Skip to main content

Full text of "ERIC ED459355: A Comparative Analysis of Transitions from Education to Work in Europe (CATEWE). Final Report [and] Annex to the Final Report."

See other formats


ED 459 355 

CE 082 741 








Smyth, Emer; Gangl, Markus; Raffe, David; Hannan, Damian F.; 
McCoy, Selina 

A Comparative Analysis of Transitions from Education to Work 
in Europe (CATEWE) . Final Report [and] Annex to the Final 
Report . 

Economic and Social Research Inst., Dublin (Ireland) . ; 
Economic and Social Research Council, Edinburgh (Scotland) . 
Centre for Educational Sociology.; Centre d' Etudes et de 
Recherches sur les Qualifications, Marseilles (France). 
Commission of the European Communities, Brussels (Belgium) . 

776p . ; Prepared with Walter Muller, Cristina Iannelli, Karen 
Brannen, Maarten Wolbers, Michele Mansuy, Yvette Grelet, 
Thomas Couppie, Gwenaelle Thomas, Hans Rutjes, Jannes 
Hartkamp, and Lena Schroder. Other partners include: 
Mannheimer Zentrum fur Europaische Sozialf orschung ; 

Research Centre for Education and the Labour Market, 
Maastricht; DESAN Market Research, Amsterdam; and Instituet 
for Social Forskning, Stockholm. Proposal no: ERB142 
PL97/2100. Funded by the Targeted Socio-Economic Research 
program, European Commission and by Deutsche 
Forschungsgemeinschaf t . 

SOE2 -CT97 - 2019 
For full text: 

ht tp : / /www . mzes . uni -mannheim . de/pro j ekte/ catewe/publ / publ_e . 
html . 

Reports - Research (143) 

MF05/PC32 Plus Postage. 

Apprenticeships; Comparative Analysis; *Comparative 
Education; Demography; *Education Work Relationship; 
♦Educational Attainment; Educational Status Comparison; 
♦Employment Patterns; Entry Workers; Foreign Countries; 
Immigrants; International Studies; *Labor Market; Outcomes 
of Education; Postsecondary Education; Secondary Education; 
Sex Differences; Social Status; Transitional Programs; 
Unemployment; Vocational Education; * Young Adults; Youth 
♦Europe; France; Ireland; Italy; Netherlands; Scotland; 
Spain; Sweden; United. Kingdom; West Germany 


This project aimed to develop a more comprehensive 
conceptual framework of school -to -work transitions in different national 
contexts and apply this framework to the empirical analysis of transition 
processes across European countries. It drew on these two data sources: 
European Community Labor Force Survey and integrated databases on national 
school leavers' surveys in France, Ireland, the Netherlands, Scotland, and 
Sweden. Three broad types of national systems were identified: countries with 
extensive vocational training systems at upper secondary level, linked to 
occupational labor markets (Germany, the Netherlands) ; countries with more 
general education systems with weaker institutionalized linkages to the labor 
market (Ireland) ; and Southern European (SE) countries with less vocational 
specialization and lower overall attainment than the other groups. In 

Reproductions supplied by EDRS are the best that can be made 
from the original document. 

"vocational" systems, young people tended to make a smoother transition into 
the labor market, while those in SE countries found it more difficult to - 
achieve a stable employment position. Educational level was highly predictive 
of transition outcomes, which varied by gender, social class, and national 
origin. Early educational failure had serious negative consequences for young 
people across all systems. Sixty-three references are listed. A separate 
annex contains these 17 working papers: "Education and Unemployment" (Brauns, 
et al.); "Position of Young People and New Entrants in European Labor 
Markets" (Couppie, Mansuy) ; "New Entrants and Experienced Workers on European 
Labor Markets" (Couppie, Mansuy) ; "European Perspectives on Labor Market 
Entry" (Gangl) ; "Education and Labor Market Entry Across Europe over the Last 
Decade" (Gangl) ; "Changing Labor Markets and Early Career Outcomes" (Gangl) ; 
"Transition from School to Work in Southern Europe" (Iannelli) ; "Educational 
Attainment of Young People in the European Union (EU) " (Mueller, Wolbers) ; 
"Integration of Young People into the Labor Market Within the EU" (van der 
Velden, Wolbers) ; "Learning and Working" (Wolbers) ; "Transition Process" 
(Grelet , et al.); "Route to Skills" (Hartkamp, Rutjes) ; "Apprenticeship in 
Ireland, the Netherlands, and Scotland" (Hartkamp, Rutjes) ; "School Effects 
on Youth Transitions in Ireland, Scotland, and the Netherlands" (Iannelli, 
Soro-Bonmati) ; "Young Immigrants on the Labor Market in France and Sweden" 
(Mansuy, Schroeder) ; "Relative Labor Market Disadvantage Among the Least 
Qualified in Ireland, the Netherlands, and Scotland, 1979-97" (McCoy) ; and 
"Gender Differentiation in Education and Early Labor Market Transitions" 
(Smyth) . (YLB) 

Reproductions supplied by EDRS are the best that can be made 
from the original document. 




A Comparative Analysis of Transitions 
from Education to Work in Europe (CATEWE). 
Final Report 

Annex to the Final Report 

Emer Smyth, Markus Gangl, David Raffe, 
Damian F. Hannan, and Selina McCoy 

Office of Educational Research and Improvement 






□ This document has been reproduced as 
received from the person or organization 
originating it. 

E. Smyth 

□ Minor changes have been made to improve 
reproduction quality 


• Points of view or opinions stated in this 

document do not necessarily represent official 
OERI position or policy. 













Contract no. : SOE2-CT97-20 1 9 

Proposal no.: ERB142 PL97/2100 

Title: A Comparative Analysis of Transitions from Education to 

Work in Europe (CATEWE) 

Project co-ordinator: Economic and Social Research Institute (ESRI), Dublin 

Partners: Centre for Educational Sociology (CES), Edinburgh 

Mannheimer Zentrum fur Europaische Sozialforschung 

Research Centre for Education and the Labour Market (RO A), 

Centre d'Etudes et de Recherches sur les Qualifications 
(CEREQ), Marseille 
DESAN Market Research, Amsterdam 
Institutet for social forskning (SOFI), Stockholm 

Reference period: from December 1997 to January 2001 

Starting date: December 1 997 Duration: 3 years 

Date of issue: 30 January 2001 

Project financed within the TSER programme 







Off S of^du^Uon^l^se^ h 


J This document has been reproduced as 
■ received from the person or organization 
originating it. 

□ Minor changes have been made to 

improve reproduction quality. 

Points of view or opinions stated in this 
documemdN not necessarily represent 


A Comparative Analysis of 

Transitions from Education to Work in Europe (CATEWE): 

Final Report 

This report was prepared by: 

Emer Smyth (ESRI) 

Markus Gangl (MZES) 

David Raffe (CES) 

Damian F. Hannan (ESRI) 

Selina McCoy (ESRI) 

With the assistance and co-operation of: 

Walter Muller (MZES) 

Cristina Iannelli, Karen Brannen (CES) 

Maarten Wolbers (ROA) 

Michele Mansuy, Yvette Grelet, Thomas Couppie, Gwenaelle Thomas (CEREQ) 
Hans Rutjes, Jannes Hartkamp (DESAN) 

Lena Schroder (SOFI) 


Table of contents 

Abstract i 

1. Executive Summary 1 

2. Background and objectives 13 

3. Scientific description of project results and methodology 18 

4. Conclusions and policy recommendations 89 

5. Dissemination 114 

6. References 120 

Annexes (in separate volume) 

1. Project deliverables 

2. Working papers 


The two main aims of the CATEWE project were to develop a more comprehensive 
conceptual framework of school to work transitions in different national contexts, and 
apply this framework to the empirical analysis of transition processes across European 
countries. The project drew on two complementary data sources for these analyses: 
the European Community Labour Force Survey, and integrated databases based on 
national school leavers' surveys in France, Ireland, the Netherlands, Scotland and 

Transition processes and outcomes were found to vary significantly across European 
countries. Three broad types of national system were identified: countries with 
extensive vocational training systems at upper secondary level, linked to occupational 
labour markets (such as Germany and the Netherlands); countries with more general 
educational systems with weaker institutionalised linkages to the labour market (such 
as Ireland); and Southern European countries with less vocational specialisation and 
lower overall attainment than the other groups. Each of these groups had distinctive 
patterns of labour market integration among young people. In 'vocational' systems, 
young people tend to make a smoother transition into the labour market while those in 
Southern European countries find it more difficult to achieve a stable employment 

Across European countries, educational level is highly predictive of transition 
outcomes; those with lower levels of education have higher unemployment risks and 
greater chances of entering low-skilled, lower status and/or temporary jobs. Those 
who have taken part in vocational education/training (especially apprenticeships) tend 
to have a smoother transition to their first job and achieve more stable employment. 
Other dimensions of education are also significant with examination grades having a 
greater effect in more general education systems. Transition outcomes are found to 
vary by gender, social class background and national origin. There is no evidence that 
such differences have become less important in shaping the transition process over 

Given the diversity in education, training and labour market systems across Europe, 
the same policy interventions are unlikely to be equally effective in different contexts. 
However, early educational failure has serious negative consequences for young 
people across all systems. There is a need, therefore, for policy intervention to reduce 
such failure and/or to provide alternative routes to skill acquisition for young people. 
There is also a need to monitor differences among groups of young people in terms of 
gender, social class and ethnicity and pursue policies to address these inequalities. 

The project highlighted a number of areas which should be prioritised in future 
research: the role of field of education/training in transitions, employer recruitment 
strategies in relation to young people; young people's own views of the transition 
process; the role of policy interventions (especially youth programmes); and 
regional/local differences in educational and transition outcomes. It is recommended 
that a European-wide survey should be initiated, covering young people from around 
the age of fifteen and following them over a ten-year period. In the absence of such a 
survey, we recommend that the Commission should encourage agreement on a 'best 
practice' template to facilitate the partial harmonisation of existing transition surveys, 
and use of, and access to, the transitions module of the Labour Force Survey should 
be enhanced. 



Recent decades have seen rapid educational expansion and labour market changes 
across European countries. Such changes have had the greatest impact on those 
entering the labour market for the first time. Indeed, many commentators have argued 
that the period of transition from school to work has become more prolonged and less 
predictable as a result of such changes. It is, therefore, crucial that we understand the 
way in which the education, training and labour market systems interact to shape the 
transition process in modem Europe. 

In spite of the importance of the transition issue, existing research has not yielded an 
adequate understanding of the processes at work across European countries. Cross- 
national studies have often focused on a narrow range of countries and have 
frequently neglected important dimensions of the transition from school to work. The 
CATEWE project set out to fill this gap by developing a more adequate conceptual 
framework to examine the relationship between education, training and labour market 
systems in different national contexts, and by applying this framework to empirical 
studies of transition processes in several European countries. 

The original objectives of the project were: 

1. To develop a more adequate and comprehensive conceptual framework, drawing 
on existing research and new analyses, of: 

• the nature of national systems of initial education and vocational training in 
European societies; 

• the factors and processes affecting variation in the full range of education/ 
training outcomes by different groups of young people in each system; 

• the processes of transition from initial education/training to work, the main 
outcomes of such transitions and pathways, and the main factors affecting 
success and failure in such transitions in each system; 

• the impact of national institutional differences on education/training outcomes 
and transition processes among young people. 

2. Using this conceptual framework, to construct an integrated cross-national data set 
using national school leavers' surveys for France, Ireland, the Netherlands and 
Scotland (UK), countries with widely varying education/training systems and 
labour market structures. 

3. To analyse education to work transitions across all European countries using the 
Labour Force Surveys, placing the analysis of school leaver transitions in the 
context of the wider European context. 

4. Using these comprehensive datasets, 

• To test and refine the conceptual framework to develop a more adequate and 
comprehensive framework to study school to work transitions across all 
European countries; 

• To explore national similarities and differences in ET systems and their 
outcomes, at an aggregate level and for different groups of young people; and 
the way in which national differences in these respects are influenced by 
institutional factors; 

• To identify the main factors influencing success or failure in ET outcomes and 
labour market integration in each system; and attempt to explain similarities and 
differences in these patterns across the different national systems. 

5. To develop proposals to harmonise existing school leavers' surveys in the 
participating countries; and encourage the extension of comparative transition 
surveys to other European countries currently planning surveys of school leavers. 

In the course of the project, these objectives were broadened in a number of ways. 
Firstly, analyses of the Eurostat Labour Force survey indicated that the Southern 
European countries had a distinctive profile in terms of labour market entry, an area 
of research that had hitherto been neglected. For this reason, it was decided to further 
explore the specific situation of young people in the Mediterranean countries by 
analysing Labour Force Survey microdata for Spain and Italy. Secondly, in order to 
broaden the number of countries included in the school leavers' survey database, 

contacts were established with a Swedish partner to utilise the Swedish cohort survey 
of young people. The inclusion of Sweden represented a substantive addition to the 
project since it has a number of distinctive characteristics in terms of the transition 
process, including a long-term decline in inequality of educational opportunity (on the 
basis of socio-economic background), the absence of a formal apprenticeship system 
coupled with extensive provision of youth programmes. Further exploration of the 
availability and coverage of school leavers' survey data indicated the value of going 
beyond the original parameters of the project to construct comparative databases not 
only for a recent time-point but to allow us to study trends over time as well as young 
people's experiences over their first five years in the labour market. Due to our 
concern with contributing to the debate on data harmonisation, the CATEWE team 
initiated co-operation with two groups (in Flanders and Portugal) who were planning 
national transition surveys. 


The initial conceptual framework identified three sets of dimensions necessary to 
explore transitions in comparative perspective: the demographic, economic and labour 
market context within which transitions occur; the dimensions of the education and 
training system; and the nature of transition processes and outcomes. 

The context within which the transition process has often been neglected in 
comparative studies of labour market entry. However, important contextual 
differences are evident among European countries. Firstly, they differ not only in the 
industrial and occupational structure of employment but also in relation to the type of 
labour market structuration (particularly whether occupational or internal labour 
markets have greater importance). Significant variation is also evident in the extent of 
labour market regulation which is likely to affect the ease of access of young people 
to (stable) employment. Macroeconomic conditions, including the nature of economic 
development and the stage in the economic cycle, are also likely to have a 
disproportionate impact on young people's labour force situation. Other contextual 
factors which need to be considered include the nature of the family- and State-based 
welfare provision, and the age, gender and ethnic structuring of the labour force as a 




Previous comparative research on transitions tended to focus on two aspects of 
education/training systems, standardisation (that is, the extent to which curricula, 
assessment and certification are nationally or regionally standardised) and track 
differentiation (the division between vocational and academic/general tracks). The 
CATEWE project went beyond these approaches to examine a broader range of 
dimensions. In particular, we were concerned with examining the way in which 
education/training systems 'sort' or differentiate their students not just in terms of 
programmes within a stage (for example, between vocational or academic) but also in 
terms of progression into the next stage and outcomes at the end of a stage (such as 
examination grades). We also considered the nature of school to work linkages, that 
is, the role of employers in the education/training system, as an important dimension. 
Our framework did not just focus on school-based education but also explored cross- 
national variation in the nature of youth training (through apprenticeships and youth 
programmes), along with its linkage to initial education and the labour market. 

These different dimensions of the education/training system can be seen to interact 
with contextual features to produce a 'transition system': the relatively enduring 
features of a country's institutional and structural arrangements which shape transition 
processes and outcomes. The important features of the transition process were seen to 
include the number and sequence of transitions and the nature of the resulting 
trajectories, the length of the transition period (from leaving education to 'stable' 
employment, for example), differentiation between transition statuses (for example, 
between 'employment' and youth training), and inequalities in terms of gender, social 
class and ethnicity. A wide range of transition outcomes were considered including: 
principal economic status; occupational status; industrial allocation; labour market 
segment; pay; and access to training among young people. 

While conceptually distinct, different dimensions of the framework interact to 
produce clusters of national systems. A broad continuum of European systems is 
evident, ranging from countries with high standardisation, strong track differentation 
and strong linkages between education and the labour market (for example, Germany 
and the Netherlands) to countries with equally high standardisation but much weaker 
track differentiation and school to work linkages (Ireland and Scotland, albeit with 

strong market signals in terms of educational qualifications). However, additional 
features which may cross-cut this continuum must be considered, including the 
strength of labour market regulation within a national system, along with the nature of 
the formal and informal (primarily family-based) welfare regimes, a feature which has 
particular relevance to Mediterranean countries. Finally, it should also be noted that 
different parts of an education/training system may have different characteristics (for 
example, school-based provision may differ markedly in nature from post-school 
vocational training). 


The CATEWE project drew on two data sources with complementary strengths for 
the analysis of school to work transitions: the Eurostat Labour Force Survey (EULFS) 
and integrated databases based on national school leavers' surveys (SLS). 

The EULFS is a cross-sectional survey, covering all EU member states, and is 
constructed to a comparable framework in order to facilitate cross-national 
comparison. As such, it enables us to study the full diversity of national contexts 
across Europe and to compare the situation of young people with that of their adult 
counterparts. It also contains detailed information on employment outcomes, allowing 
us to examine their relationship with level and type of education. However, the cross- 
sectional nature of the survey means that we cannot directly examine the transition 
process itself. Instead, synthetic labour force entry cohorts (based on 'typical' age of 
graduation from different levels and types of education) have been constructed to 
examine patterns across different cohorts. 

A set of integrated databases were constructed drawing on national transition surveys 
in France, Ireland, the Netherlands, Scotland and Sweden. Key measures of education 
and transition outcomes were available in these national surveys and they were 
mapped to a common template of comparable variables, a process which also helped 
us to refine the conceptual framework described above. The advantage of this data 
source is that national school leavers' survey explicitly take account of the transition 
process by examining the flow out of initial education into the labour market or 
further education/training and allow us to observe individual pathways rather than the 

aggregate patterns possible with the EULFS. These surveys also contain detailed 
information on the main dimensions of education/training outcomes, enabling us to 
explore their relative importance in different institutional contexts. However, the use 
of integrated databases is somewhat limited in terms of the number of European 
countries covered, the absence of information on older age-groups for the purposes of 
comparison, and some difficulties in comparability (for example, the absence of 
family background information from some national surveys). 

The following section highlights the main findings of the LFS and SLS analyses. 
These are presented separately for the purposes of clarity but section 1.5 will draw 
together the two sets of findings in discussing the implications of the CATEWE 
project for policy and future research. 


The CATEWE project resulted in twenty-two substantive papers on different issues 
relating to school to work transitions, including the nature of the relationship between 
education and labour market outcomes, the nature of apprenticeship systems, variation 
in youth programme provision, and differentiation among groups of young people in 
terms of gender, social background and national origin. These papers draw on 
complex multivariate analyses which enable us to compare European countries in a 
systematic way and allow us to highlight the significant dimensions of education and 
training influencing the transition process. The remainder of this section presents a 
broad overview of the main project findings. 

1.4.1 The Eurostat Labour Force Survey 

Analyses of the Eurostat Labour Force Survey indicate some convergence in levels of 
educational attainment across Europe with the greatest increases evident in countries 
with previously low levels of attainment. However, there is no evidence of any 
convergence in the type of education across countries. In this respect, analyses 
identified three ideal-typical groups of countries within Europe: countries with an 
extensive system of vocational training at upper secondary level (either 
apprenticeship-based as in Germany or school-based as in the Netherlands) and 

consequently a high prevalence of labour market entrants with occupationally-specific 
qualifications; Northern European countries with fairly large proportions of entrants 
having general rather than vocational qualifications; and Southern European countries 
with lower levels of educational attainment overall and less vocational specialisation. 

These three groups of countries were found to have quite distinct patterns of labour 
market integration among young people. Across all European countries, labour market 
entrants tend to experience higher unemployment rates, more employment instability 
and have lower skilled jobs than more experienced workers. However, the gap 
between these groups is more pronounced in Southern European countries with 
entrants being vulnerable to proportionately higher unemployment rates and more 
precarious employment. Northern European countries were found to occupy an 
intermediate position, though with quite a distinction evident between 'outsiders' 
(labour market entrants) and 'insiders' (more experienced workers) particularly in 
France, for example. In contrast, unemployment rates among labour market entrants 
closely parallel those among more experienced workers in systems with a strong 
emphasis on the provision of occupationally-specific skills, such as Austria, Germany, 
Denmark and the Netherlands. 

Among labour market entrants, educational level is found to be highly predictive of 
transition outcomes. Those with lower levels of education have significantly higher 
unemployment risks than the better qualified across most European countries, with the 
exception of Southern Europe. Furthermore, educational level is also associated with 
the quality of employment secured. Those with higher educational levels tend to 
achieve higher occupational status, have a lower likelihood of entering low-skilled or 
temporary jobs, an enhanced access to professional positions and are more likely to 
obtain full-time contracts. Type of education is also crucial, although its role varies 
across different institutional contexts. Those who have taken part in vocational 
education/training (especially apprenticeships) tend to have a smoother transition to 
their first job and also tend to access more stable employment. In countries without 
extensive upper secondary vocational training, post-school training tends to reinforce, 
rather than compensate for, initial levels of education. 





Among labour market entrants, unemployment risks are also related to aggregate 
macroeconomic conditions with the least qualified being particularly vulnerable to 
cyclical swings. Macroeconomic conditions have a much stronger influence on 
unemployment risks than they do on the quality of the job obtained. However, there is 
evidence of some changes in job quality over time; educational expansion is 
associated with lower net returns to education in terms of occupational status and skill 
level. This is, at least partly, counterbalanced by the tendency of educational 
expansion to be associated with an increasing professionalisation of the labour force. 

1.4.2 Analyses of SLS data 

Analyses of school leavers' survey data focused on a narrower range of countries but 
the availability of a wider range of variables and multidimensional indicators of 
educational and transition outcomes enabled us to explore heterogeneity among this 
(mainly Northern European) group. As with analyses of the Labour Force Survey, 
level of education was found to be highly predictive of transition outcomes among 
young people. Those who leave school with lower levels of education have a higher 
risk of being unemployed (immediately after leaving school and over the first five 
years in the labour market) and their unemployment spells tend to be longer in 
duration. Some cross-national variation is evident in the distribution of unemployment 
with unqualified school-leavers in Ireland experiencing more long-term 
unemployment while in France unemployment tends to be interspersed with periods 
of participation in youth programmes or short-term employment. When they secure 
employment, the least qualified tend to enter part-time, lower status and/or lower 
skilled jobs. 

Other dimensions of educational differentiation are found to influence transition 
outcomes, but often in different ways in different institutional systems. Examination 
grades are found to have a more significant effect on transition outcomes in more 
general education systems than in more track-differentiated systems. Higher- 
performing students have reduced unemployment risks in Ireland, Scotland and 
Sweden but grades are not significantly associated with unemployment chances in the 
Netherlands where type and level of education play a more important role. 

School-based vocational education has the strongest labour market effects in formally 
track-differentiated systems with occupational ised labour markets (such as the 
Netherlands). However, there is some evidence (as in the Irish case) that the 
development of occupationally-specific vocational courses within the framework of a 
'general' education system may yield similar benefits. Post-school vocational 
education in the form of apprenticeship and participation in youth programmes was 
also considered. The apprenticeship system plays a distinctive role across countries, 
forming an alternative to school-based vocational education as a route to skills in 
France and the Netherlands but operating as a type of post-school vocational training 
in Ireland and Scotland. The prevalence of youth programmes also varies across 
countries, being higher in Scotland, France and Sweden than in Ireland or the 
Netherlands. In part, this is related to greater labour market regulation in France and 
Sweden with schemes providing work experience as a means of labour market access. 
However, in Scotland youth programmes have become an important route to skill 
acquisition even in the context of a relatively 'flexible' labour market. 

The nature of the education/training system influences the extent of differentiation in 
terms of gender and ethnicity. In particular, early selection into different tracks plays 
a role in increasing differentiation in terms of educational and transition outcomes. In 
the Netherlands, for example, gender differences in the type of vocational tracks taken 
is associated with somewhat higher occupational and industrial segregation by gender 
on entry to the labour market. However, it should be noted that gender segregation 
within the labour market was a feature of all of the countries considered. In a similar 
way, early selection in the French system is associated with greater differences in 
educational outcomes between immigrant and native-born young people. Analyses 
from the CATEWE project indicate no evidence that educational and transition 
outcomes have become less structured over time by background characteristics such 
as gender, social class and ethnicity. 



Perhaps the main conclusion of the CATEWE project is that, given the diversity in 
education, training and labour market systems across Europe, the same policy 
interventions are unlikely to be equally effective in different contexts. It is worth 



noting, however, that, while relatively enduring in the way they shape school to work 
transitions, national transition systems are not fixed. Indeed, many of the countries 
considered in our analyses have experienced considerable changes in recent decades, 
particularly in the nature of their apprenticeship and youth programme provision. 
Transition processes and outcomes are, therefore, amenable to policy intervention but 
not to the imposition of a single 'solution' derived from a very different institutional 
context. It has proved difficult, for instance, to expand the apprenticeship system 
beyond traditional crafts sectors in certain systems (such as Ireland) so this would not 
appear to be a viable solution to early educational failure in all contexts. 

A fairly striking regularity across Europe is the crucial role of educational level in 
shaping transition outcomes. Those with low levels of education/training continue to 
experience marginalisation within the labour market (either through increased 
unemployment or precarious work situations), even in the context of rapid 
employment growth. There is, therefore, a need for policy intervention to prevent 
drop-out from the education system and/or to provide alternative routes to skills 
acquisition and labour market integration for young people who have experienced 
educational failure. 

The provision of vocational education may be seen as one means of achieving this 
end. All else being equal, young people who have taken a vocational track are found 
to have a smoother transition to the labour market and tend to access better 
employment opportunities. This pattern is particularly marked for those who have 
taken an apprenticeship programme. But the potential disadvantages of promoting 
greater commitment to vocational education must also be considered. Young people 
who have taken vocational tracks are less likely than those who have taken general 
tracks to enter higher education and, in the longer term, they tend to be excluded from 
higher status occupations. Early selection into different educational tracks may also 
exacerbate social differences in educational outcomes, leading to more unequal 
outcomes for working-class and ethnic minority youth. 

Policy interventions thus need not only to respond to 'average' transition patterns but 
to take account of diversity among young people in terms of gender, social class and 
ethnicity. Discussions about gender equity have often focused on the position of 

women 'returners' to the labour force. However, our analyses indicate the persistence 
of segregation by gender among young women in the face of the introduction of equal 
opportunity legislation across Europe. More research is, therefore, needed to examine 
both the formal and the informal factors shaping choices on the part of employers, 
education/training providers and young people. There is also a need for more 
information on the social class and ethnic background of young people and the 
development of equity measures in education and transition outcomes. 

Other research is needed to inform policy in a range of areas, including an 
investigation of: 

• employer strategies in different segments of the labour market, in particular, their 
decisions in relation to the recruitment and training of young people; 

• young people's own views of the transition process, their expectations and 

• the role of policy interventions, especially youth programmes, in the transition 
process and their relationship with other forms of education and training; 

• the role of field of education/training in shaping transition outcomes, controlling 
for level of education; 

• regional and local differences in educational and transition outcomes. 

It is recommended that the potential use of existing data sources should be enhanced 
in order to study transition processes from a comparative perspective. Firstly, 
improved documentation of, and access to, EULFS microdata would facilitate 
transitions research. Secondly, the transitions module added to the Labour Force 
Survey has considerable potential, but only if researchers are granted access to the 
microdata. In addition, the potential of the module to collect information on social 
background and career trajectories should be enhanced in future waves of the survey. 
Thirdly, while full harmonisation of existing national transition surveys is not 
feasible, it is recommended that agreement should be reached on a template which 
represents best practice and principles for the partial harmonisation of these surveys. 
The added value of attempting such harmonisation could be high, particularly as the 
number of European countries conducting such surveys has been increasing. Finally, 
we recommend the initiation of a European-wide survey based on a prospective age 





cohort design starting at about fifteen years, followed over a period of about ten years. 
This would enable us to examine decision-making processes among young people at 
the point of leaving compulsory education and their subsequent trajectories through 
the education and labour market systems. The improvement of existing data sources 
coupled with the collection of new data would greatly enhance our ability to 
understand transition systems across Europe in years to come. 


A number of complementary dissemination strategies were adopted within the life- 
time of the CATEWE project. These included the presentation of conference papers, 
publication in journal and book format, the creation of a CATEWE web-site 
containing working papers ( '), and 
the development of links with OECD and other relevant organisations in relation to 
proposals for the harmonisation of transitions data. Future activities will include the 
publication of a book on the Labour Force Survey analyses, the publication of papers 
on the School Leaver Survey analyses in scientific journals, along with the use of 
existing networks among project partners to carry out dissemination to policy-makers 
and other interested parties at a national level. 


In recent decades, European countries have experienced significant growth in the 
educational qualifications of labour market entrants. At the same time, occupational 
structures and employment practices have undergone considerable change across 
Europe. Such changes highlight the necessity of understanding the relationship 
between the education/training system and labour market changes in different national 
contexts. In particular, the persistence of youth unemployment in many European 
countries means that attention should be given to the period of early labour market 
integration. It is also important that the transition from education to the labour market 
should be examined in comparative perspective. The diversity of institutional 
arrangements across EU member states means that we cannot assume that policy 
interventions will operate in the same way in different contexts. Instead, we need to 
identify the important dimensions of national education, training and labour market 
systems in order to formulate effective policy in this regard. 

A number of researchers have examined institutional variation in education, training 
and labour market systems, proposing that the nature of the relationship between these 
elements is societally specific (see, for example, Maurice et al., 1986). While such 
work has improved our understanding of the transition process, empirical research has 
proved rather limited and narrowly focused. Cross-national studies have tended to 
focus only on two or three countries, running the risk of generalising to very different 
institutional contexts on the basis of a limited number of 'core' European countries 
(usually Germany, France and/or Britain). Measures of education and labour market 
outcomes have often obscured very real differences between systems and many 
crucial dimensions of the transition process have been ignored. The CATEWE project 
set out to address this deficit by developing a more adequate and comprehensive 
conceptual framework for examining the relationship between education, training and 
labour market systems, and applying this framework to a comparative empirical study 
of recent developments in several European countries (see Hannan et al., 1999). 

13 19 


The project developed out of the experiences of project partners in participating in the 
European Network on Transitions in Youth. This Network was established in 1992, 
drawing together researchers who had considerable experience in conducting and 
analysing national transition surveys as well as researchers who had used Labour 
Force Survey data to analyse the relationship between educational and labour market 
outcomes. Participation in the Network increased awareness of national differences in 
transition processes as well as of the kind of information available to study transitions 
and stimulated a number of small (usually two-country) comparative studies'. It 
quickly became clear that it was necessary to expand the range of countries 
considered in order to capture the most important dimensions of institutional 

Discussions around this topic led to the preparation of a paper for the OECD, which 
was then preparing a Thematic Review of the Transition from Initial Education to 
Working Life, outlining a preliminary framework for analysing institutional 
differences in school to work transitions (Hannan, Raffe, Smyth, 1997), and to the 
formulation of a research project funded under the Leonardo project. This project 
drew on national transition surveys to focus on the experience of 'lower level leavers' 
(those leaving school without upper secondary certification) in four countries: France, 
Ireland, the Netherlands and Scotland (see Hannan, Smyth et al., 1998). The 
experience of attempting to develop an integrated cross-national database highlighted 
the potential, as well as the difficulties, involved in such an exercise. It also became 
clear that an adequate analysis of institutional diversity would need to take account of 
a broader range of national systems. As a result, the CATEWE project aimed to draw 
on two complementary sources of data: the Eurostat Labour Force Survey and 
national school leavers' surveys. 

1 Proceedings of the European Network on Transitions in Youth annual workshops have been 
published in CEDEFOP (1994), ESRI, Combat Poverty Agency (1998), Raffe et al. (1999), and 
Hammer (2000). 


The original objectives of the project were: 

1. To develop a more adequate and comprehensive conceptual framework, drawing 
on existing research and new analyses, of: 

• the nature of national systems of initial education and vocational training in 
European societies, with particular reference to their degree of differentiation 
and standardisation as well as to the relationship between general and 
vocational education in each system; 

• the factors and processes affecting variation in the full range of education/ 
training outcomes by different groups of young people in each system; 

• the processes of transition from initial education/training to work, the main 
outcomes of such transitions and pathways, and the main factors affecting 
success and failure in such transitions in each system; 

• the impact of national institutional differences on education/ training outcomes 
and transition processes among young people. 

2. Using this conceptual framework, to construct an integrated cross-national data set 
using national school leavers' surveys for France, Ireland, the Netherlands and 
Scotland (UK), countries with widely varying education/training systems and 
labour market structures. Strenuous efforts were to be made to incorporate other 
countries who had conducted, or were planning to conduct, school leavers’ 

3. To analyse education to work transitions across all European countries using the 
Labour Force Surveys, placing the four country analysis of school leaver 
transitions (2) in the context of the wider European context. 

4. Using these comprehensive datasets, 

• To test and refine the conceptual framework (1) and associated hypotheses; and 
to develop a more adequate and comprehensive framework to study school to 
work transitions across all European countries; 

• To explore national similarities and differences in ET systems and their 
outcomes, at an aggregate level and for different groups of young people; and 
the way in which national differences in these respects are influenced by 
institutional factors; 

• To identify the main factors influencing success or failure in ET outcomes and 
labour market integration in each system; and attempt to explain similarities and 
differences in these patterns across the different national systems. 

5. To develop proposals to harmonise existing school leavers' surveys in the 
participating countries; and encourage the extension of comparative transition 
surveys to other European countries currently planning surveys of school leavers. 

It was intended that analyses should be framed in terms of five key research 


• What is the nature and extent of similarities and differences in education/training 
systems within the EU countries studied, and in the associated type and level of 
education and training achieved by educational system leavers entering the labour 

• What is the relationship between differences in education/training outcomes and 
the social background characteristics of system leavers (in terms of gender, social 
class and ethnic origin)? And do such differences vary systematically across 
national systems? 

• How do transition processes vary systematically across countries (in terms of their 

length, complexity etc.)? And to what extent are these differences related to 
differences in education, training and labour market structures? 

• What is the nature and extent of the relationship between level and type of 
educational achievements among system leavers and (the success of) their 
transition processes and outcomes? How do these relationships vary by type of 

• What is the relationship between social background characteristics and labour 
market outcomes? To what extent is this relationship mediated by education, and 
does this vary by type of system? 




The following section outlines how these objectives were met, and research questions 
addressed, during the course of the project. However, the objectives themselves 
became subject to review and development over the course of the project. Firstly, it 
became clear that, at least in terms of national transition survey data, information was 
available only on a limited range of European countries. As a result, an attempt was 
made to broaden the scope of countries included. Information from a Swedish cohort 
study was incorporated into the school leavers' database(s) and links were forged with 
two countries (Belgium (Flanders) and Portugal) who were initiating transition 
surveys. Furthermore, analyses of the Labour Force Survey data indicated the 
distinctive character of institutional and market systems in Southern Europe. For this 
reason, analyses of microdata for Italy and Spain were conducted in order to further 
explore the source of such variation. Secondly, the construction of an integrated 
database using school leavers' survey data revealed the rich potential of this data 
source. As a result, four separate integrated datasets were constructed, allowing us to 
examine cross-national differences at one point in time, over the 1980s and 1990s, 
and five years after leaving school. Further details on the development of the project 
are presented in the remainder of the report. 







In this chapter we describe the main activities of the project and summarise its findings. We 
start, in section 3.2, by outlining the project’s conceptual framework, paying most attention to 
dimensions of variation in national ‘transition systems’. In section 3.3 we then contrast 
different approaches to comparative research, and describe the core strategy of the project 
which involves what we describe as relatively ‘intensive’ comparisons between countries. 
The project took advantage of the complementary opportunities provided by its two main data 
sources, the Eurostat Labour Force Survey (EULFS) and national School Leavers' Surveys 
(SLS), and these are described in the following three sections. In section 3.4 we discuss the 
relative strengths and weaknesses of these two data sources. In section 3.5 we summarise our 
work on the EULFS, presenting our findings in terms of four main themes. In section 3.6 we 
similarly summarise our work on school-leaver surveys. Finally, in section 3.7 we describe 
the project’s work on new data-collection: its support for new transition surveys in two 
European countries and the development of proposals for future comparative data on 
transition, but on this and other issues we save our main recommendations for chapter 4. 


The research questions outlined in the previous chapter can be summarised as a single general 
question: “How do national transition systems shape transition processes and outcomes?” To 
answer this question we needed concepts of transition system, and of the dimensions of 
variation in national systems; and we needed an understanding of the main transition 
processes and outcomes. 

The project’s first report was submitted in summer 1998 and published as a Working Paper in 
the following year (Hannan et al., 1999). It presented the project’s initial conceptual 
framework, which was to inform the construction of datasets and the analysis of data. The 
framework built on previous research, including the review by Hannan, Raffe and Smyth 
(1997) described in chapter 2, and on country studies which summarised existing research on 



education-to-work transitions in France, Germany, Ireland, The Netherlands, Portugal and 
Scotland. The country studies were published at the same time, together with an analysis of 
the changing economic and demographic context of transition. 

The report stressed that the conceptual framework was an initial statement, for subsequent 
review and refinement, and it continued to develop during the life of the project. In the 
presentation that follows we focus on the conceptual framework as we used it in our work. 
We take the initial framework, presented in our first report, as the starting point and outline 
the main lines of development, giving most emphasis to features which proved important in 
our empirical analyses. 

The initial conceptual framework identified three elements (or groups of dimensions) in 
cross-national comparisons of transition: 

• the demographic, economic and labour-market context within which transitions occur; 

• the dimensions of the education/training system; 

• the nature of transition processes and outcomes. 

As the project developed we found it useful to combine parts of the first two elements to 
develop the concept of ‘transition system’. This embraced the main independent variables of 
our analyses: the macro-level ‘determinants’ of transition patterns. The third element of our 
framework provided the main dependent variables, the processes and outcomes of transitions 
at the micro level. The concept of ‘transition system’ describes the relatively enduring 
features of a country’s institutional and structural arrangements which shaped transition 
processes and outcomes. Our work paralleled that of the OECD’s Thematic Review of the 
Transition from Initial Education to Working Life in developing understandings of transition 
systems (OECD, 2000). The paper by Hannan et al. (1997) was written for the OECD 
Review, and members of the project and the OECD had a joint meeting in 1998. 

3.2.1 Dimensions of transition systems and their contexts 

With respect to the demographic, economic and labour-market context (the first of the three 
elements listed above), the report identified the following key dimensions: 

• the ratio of young people to adults, both in the population as a whole and in the labour 
force, including the effect of migration; 





• the nature and resources of the family system, and how this provides opportunities or 
encouragement to young people to stay at home or to set up independent households; 

• the nature of economic development: this embraces the distinction between the economic 
core and periphery of Europe, and variations in the structure and ownership of firms; 

• the stage in the economic cycle; 

• the industrial and occupational structure of employment; 

• the nature of labour-market segmentation within a country, with particular reference to the 
relative importance of occupational labour markets (OLMs), internal labour markets 
(ILMs) and external labour markets (ELMs). These distinctions are complex and countries 
constitute varying mixtures of all three types; 

• the size and nature of the informal economy; 

• the extent of regulation of the labour market; 

• the age structuring of the labour force: that is, the extent of segregation between youth and 
adult labour markets; and 

• the gender and ethnic structuring of the labour force, and the strength of segregation 
associated with these traits. 

The project focused especially on the dimensions of labour-market structure, including the 
nature of segmentation (the OLM/ILM distinction) and the extent of regulation. Together 
with features of education and training systems, described below, these were the main 
elements of the concept of transition system as it was employed in the project’s analyses. The 
ILM/OLM distinction is central to a variety of distinct but related theoretical approaches, 
including Maurice, Sellier and Silvestre’s (1986) distinction between qualification space and 
organisational space, Marsden’s (1986) analysis of labour-market segmentation and Garonna 
and Ryan’s (1991) analysis of modes of inclusion and exclusion in youth labour markets. 

Many of the dimensions listed above are of interest more on account of their variation over 
time, than of their variation between countries. For example, transition processes and 
outcomes are influenced by the stage of the economic cycle and by trends in the occupational 
or industrial structure of employment. 

With respect to the second element, the dimensions of education and training systems, the 
project’s first report identified the following dimensions or sets of dimensions: 

• standardisation: the extent to which curricula, assessment and certification, and related 
quality assurance procedures, are standardised on a national or regional basis. 
Standardisation may vary across different parts of the same system: for example 
vocational qualifications may be more standardised than general qualifications, or vice 
versa. However the project made less use of the concept of standardisation in its empirical 
analyses, partly because of the limited variation across European countries; 

• differentiation: the extent and manner in which a system differentiates its students. The 
initial conceptual framework noted that this might differ between stages of education 
within the same system. It further distinguished three types of differentiation, which 
respectively concerned 

♦ institutions or programmes within a stage, 

♦ progression into the next stage, and 

♦ outcomes at the end of a stage, especially the level of attainment or grades 

The project’s empirical analyses confirm the multi-dimensional nature of differentiation, 
but do not settle on a single way to represent these dimensions. Many analyses work with 
a simpler concept of track differentiation, which broadly corresponds to the first two types 
of differentiation listed above, as applied to Tower’ and ‘upper’ secondary education and 
the relation between them. However some systems with relatively low levels of track 
differentiation, such as Ireland or Scotland, may have strong vertical differentiation with 
respect to attainment at the end of each stage, so the project sometimes found it useful to 
distinguish outcome differentiation (attainment at the end of a stage) from track 

• school-to-work linkages: the role of employers in the education/training system. At one 
pole are systems where employers are direct providers of education/training, for example 
through apprenticeship. Other systems (such as the Netherlands) have collinear 
relationships, with employers making an important institutionalised input into school- 
based vocational education. A third category (such as Japan) comprises systems where 
employers have links with schools for recruitment. At the other pole are systems where 
employers have little direct involvement in any of the senses described above. These 
comprise two sub-categories: systems in which recruitment decisions are substantially 
based on individuals’ educational attainments, which therefore send out strong market 
signals to education; and systems where this does not happen and market signals are 




weak. Different types of linkage may be found in different sectors of the same education 
and training system, but there is cross-national variation in the linkages that are 
characteristic of equivalent sectors; 

• youth training: arrangements for youth training, and work-based provision more 
generally, vary with respect to the level of provision, the degree to which it is 
differentiated (including between apprenticeship and other programmes) and the formal 
inclusiveness of provision; 

• in addition, several CATEWE analyses have examined the level of participation and/or 
attainment in education, with particular reference to the relative scale of lower-secondary, 
upper-secondary and tertiary education, or their ISCED equivalents. 

These dimensions are neither logically nor empirically independent of each other. We discuss 
their interconnections in the next section. 

3.2.2 Classifying transition systems 

The project’s first report drew on the country studies to produce a provisional classification of 
European countries in terms of the dimensions of the conceptual framework (Hannan et al., 
1999, Figure A. 1). This classification is detailed and we do not reproduce it here, but it draws 
attention to the importance of variation within each system with respect to many of the 
dimensions examined. For example, features such as standardisation, differentiation or 
linkages with the labour market may vary between different educational tracks, as well as 
between different stages of education, within the same system. However this classification 
was too detailed to form the main basis of the project’s empirical comparisons, and it was 
necessary to develop a simpler classification with which to work. 

When countries are classified according to a matrix which combines many of these 
dimensions (see Figure 3.1) they form a broad continuum, from countries with high levels of 
standardisation, strong track differentiation, strong linkages and significant apprenticeship 
systems, to countries with weaker track differentiation (but often strong vertical ‘outcome’ 
differentiation), weak linkages and in some cases extensive youth programmes. This 
continuum is also associated with some of the labour-market dimensions discussed above, and 
especially with the distinction between OLMs and ILMs. OLM countries tend to have 




standardised, track-differentiated education systems, strong linkages and high levels of 
apprenticeship, and they therefore appear at the former end of the continuum of transition 
systems described above. 

Of the countries covered by the CATEWE country studies, Germany followed by the 
Netherlands is at one end of the continuum with strong OLMs, track differentiation and 
linkages, while the others (France, Ireland, Portugal and Scotland) are towards the latter end 
but in the category with strong market signals from the labour market to education. All are 
relatively standardised. However the project’s first report acknowledged that the distinctive 
features of the transition system of Portugal, as well as other southern European countries, 
were not fully captured by this framework, and identified this as an area for further study. In 
addition, a study would need to include non-European countries, such as the United States 
and Japan, in order to cover the full range of variation in transition systems expressed by the 
conceptual framework. 

Another important dimension of variation in national labour markets, the strength of labour- 
market regulation or flexibility, partly cuts across this continuum. The report suggested that 
this could be measured by the extent of state regulation of employment standards and 
employment protection, and by the extent of corporate coverage of trade union/employer 
relationships, although later studies refined these criteria. The strength of labour-market 
regulation differentiates among the large block of ‘loosely coupled’ countries with strong 
labour market signals. Of the main countries studied by the project, Scotland and (to a lesser 
extent) Ireland have weaker regulation than the others. 

Different CATEWE analyses emphasise different dimensions of variation in transition 
systems. This depends partly on their theoretical starting point, and partly on the data source 
and the opportunities that it provides. Analyses of EULFS data tend to focus on dimensions of 
labour-market variation, in particular the ILM/OLM distinction, rather than on variation in 
education and training systems. The EULFS provides good labour-market outcome measures 
with which to test the ILM/OLM distinction and to derive empirically-based clusters of 
countries. The EULFS analyses also distinguished countries in terms of the strength of 
labour-market regulation, and time-varying features of transition systems or their contexts 
such as levels of educational participation, occupational composition and economic activity. 

Conversely, analyses of SLS data tend to focus on the dimensions of variation in education 
and training systems. The main dimensions of transition systems investigated by SLS 
analyses are track differentiation, linkages and the role of work-based provision. 

A theoretical challenge in the analysis of transition systems is to clarify the connection 
between their educational and labour market dimensions. For example, is the connection 
between dual systems and OLMs a necessary connection, explicable in terms of the 
theoretical frameworks used to analyse labour markets and education systems, or is it merely 
a product of contingent historical circumstances? However these connections could not be 
fully explored during the project, partly because of the different emphases of the two main 
data sources, and further analysis, perhaps using different data, is required. 



Figure 3.1: A typology of ET systems and labour market linkages 

School-Work Linkage 

(a)Tightly coupled 
ET/employer systems: 
strong linkage 
(dual system) 

Degree of Standardisation of ET System 



Degree of Differentiation (and Vocat./Occupat. Specificity) of ET Systems 






► Low 



Substantial sharing and co- 
operation between providers 
and employers in delivery of 
ET. As in apprenticeships. 
High occupationalisation of 


(b) Tightly coupled 
ET/employer systems: 
collinear linkage: 


High levels of in-school provision 
of ET specific to particular 
occupations, agreed with employers. 
High occupationalisation of LM. 

(c) Loosely coupled or 
decoupled ET/employer 
systems, but with strong 
market signals: 

Low degree of ET provider and 
employer sharing of ET provision; 
low occupationalisation of LM, and 
limited school involvement in 
employment decisions. 

(d) Loosely coupled 
systems, but with strong 
market signals and 
strong school placement 










(e)De-coupled ET/LM 
systems with weak 
market signals (from 
second level). 




3.2.3 Transition processes and outcomes 

With respect to the main dependent variables of the analysis, transition processes and 
outcomes, the project’s first report notes that the process of transition has changed in most 
countries, involving a longer and more complex sequence of transitions and a larger number 
of intermediate statuses between education and a ‘stable’ status in the labour market. There is 
a debate about the extent to which youth transitions have become more individualised and 




whether this has affected the nature or importance of social and gender inequalities. Important 
features of the transition process include: 

• the number of separate transitions which comprise a ‘completed’ transition from 
education to work for an individual; 

• the length of the period over which these transitions take place; 

• the extent of differentiation between transition statuses (for example, the extent to which 
apprenticeships are distinct from other work-based training, and the extent and the nature 
of ‘dual statuses’ such as combinations of work and education); 

• the nature of trajectories: particularly the ways in which education, training, qualification 
outcomes and employment/unemployment are interrelated (for example, the phasing of 
education/training and employment status changes, the possibility of ‘reversals’ and 
‘bridging loops’ back to education/training from the labour market, and the extent of 
education/training involvement leading to qualifications); 

• inequalities in respect of gender, social class and ethnicity; and relatedly, 

• the extent of individualisation, in the sense either of a growth in the number and 
complexity of transitions, or a reduction in the correlation between transition processes 
and background characteristics such as gender and social class. 

The concept of transition outcome, and especially the definition of a successful transition, are 
problematic. In many studies the available data make it necessary to define outcomes in terms 
of ‘snapshots’ at a given point in the transition process, for example one year or five years 
after leaving school. The main ‘snapshot’ outcomes of concern to this study are: principal 
economic activity; occupational status; industrial allocation; labour market segment; wages; 
security of employment; access to on-the-job training; access to off-the-job training 
sponsored by employers; content congruence, that is, matching between type of education and 
type of occupation; ‘level congruence’, or the extent of matching between level of education 
and occupational status. However comparisons based on snapshot measures are sensitive to 
the time at which they are measured. The difference in outcomes between young people from 
different countries may be very different one year after leaving school but may become more 
similar a few years later. More fundamentally, any comparison based on a single measure is 
problematic in the context of cross-national differences in the length of the transition process, 
in the characteristic sequence of transitions and in the blurring between statuses. Other 
outcome measures try to capture aspects of the transition sequence, for example time to first 



job, percentage of time unemployed, job and career mobility, frequency of job changes or loss 
of employment; others measure more complex ‘trajectories’ of status changes over time. 
However there is a need for further conceptual development in this area. 

Finally, the project’s first report refers to various dimensions of state intervention (in military 
service requirements, youth programmes, social welfare provision as well as the labour 
market) in its discussion of transition processes and outcomes, although these more properly 
belong with the concept of transition system rather than with transition processes and 

The way in which the project operationalises transition processes and outcomes depends on 
the opportunities afforded by the data, discussed later in this chapter. Most analyses both of 
the EULFS and of the SLS use ‘snapshot’ measures of such outcomes as (un)employment or 
occupational level. The EULFS has insufficient longitudinal data, and the SLS data are 
insufficiently comparable, to permit much ambition in defining trajectories or other measures 
of the sequencing of transition statuses. Analyses of both data sources use their potential for 
correlating transition outcomes with educational background, or with gender and social 

3.2.4 Summary: conceptual framework 

To summarise, the conceptual framework developed in the project: 

• uses the concept of transition system to describe the interrelated features of education 
and training systems, national labour markets and other macro-level determinants of 
transition processes and outcomes; 

• identifies a large number of dimensions of transition systems and of their social, 
demographic and economic contexts; 

• suggests that many of these can be expressed in terms of one overarching dimension, 
associated with the strength of OLMs, apprenticeship, standardisation, track 
differentiation and education/labour-market linkages; 

• identifies other important dimensions of variation in transition systems, including the 
strength of labour-market regulation and the ‘vertical’ differentiation of levels of 
educational achievement or grades; 





• acknowledges the need for further theoretical and empirical work on the distinctive 
features of southern European transition systems; 

• acknowledges that the precise connection between the educational and labour-market 
dimensions of transition systems, and the distinction between a transition system and 
its socio-economic context, require further analysis; and 

• identifies transition processes and outcomes in several ways: in terms of ‘snapshots’ 
of individuals’ statuses at given time points, classifications of longitudinal transition 
sequences, the strength of association between education and labour-market outcomes, 
and social and gender inequalities. 


The project required comparable data on transition processes and outcomes across a sample of 
countries which represented theoretically significant variation in transition systems. We used 
two main sources of data: the Eurostat Labour Force Survey, and datasets constructed from 
national surveys of secondary school leavers in France, Ireland, the Netherlands, Scotland and 
Sweden. The project aimed to benefit from the complementary strengths of these two sources 
of data, which we discuss further (together with their respective limitations) in section 3.4. 

There are different ways in which researchers may pursue a question such as “How do 
national transition systems shape transition processes and outcomes?” Figure 3.2, based on 
Raffe (2001), summarises contrasting strategies of comparative research. At root they reflect 
the different possible purposes of comparison: do we compare countries in order to identify 
universal laws or patterns, or in order to elucidate national uniqueness? (Kohn, 1987; 0yen, 
1990; Bynner and Chisholm, 1998; Evans, 1999). A universalistic approach aims to ‘replace 
countries by variables’ (Ragin and Becker, 1992), and to identify a set of laws which not only 
transcend national differences but also explain them. It would attempt to answer our research 
question by reducing differences in national transition systems to a series of dimensions or 
types, which explain why transition processes and outcomes vary across countries without the 
need to refer to idiosyncratic national features. By contrast, particularistic research aims to 
discover the unique logic which governs social processes within each country. Each transition 
system comprises a unique set of structures, concepts and relationships which defy any 
attempt to generalise or classify across countries; even phenomena which appear to be 




general, such as entry to the labour market or the institution of apprenticeship, have different 
significance in terms of their national logics, and the task of research is to unpick differences 
between superficially similar concepts. The universalistic strategy typically involves an 
extensive approach to comparison, which uses large samples of countries in order to 
distinguish empirically among alternative country-level explanatory variables. The 
particularistic strategy typically uses a small sample of countries, often just two or three, to 
make interpretive comparisons, whose main aim is to highlight qualitative differences in 
concepts and institutions. It is this approach which underlies the view of comparative research 
as a means to gain a better understanding of one’s own country, by exposing taken-for- 
granted assumptions and opening them to challenge. 

Figure 3.2 Comparative research strategies 





To identify universal 
laws or patterns 

Mixed aims 

To elucidate national 


Replace countries with 

Use common concepts to 
describe and classify 
national logics and 
analyse differences 

Use distinctive concepts 
to describe and analyse 
internal logic of each 

Use of 


large sample of 
countries to represent 
variables of theoretical 


use small sample of 
countries; multiple 
comparisons provide 
degrees of freedom 


use small number of 
national contrasts to 
highlight differences in 
concepts and institutions 

The contrast described above is a matter of emphasis; there are few pure examples of either 
strategy. Many researchers have adopted an intermediate position (eg Maurice et al., 1986; 
Kohn, 1987), and so has the CATEWE project. In the middle column of Figure 3.2 we 
identify an intermediate strategy which may have both universalistic and particularistic 
purposes, which recognises the existence of distinctive national ‘logics’ but which tries to 
develop common cross-national concepts to describe them and cross-national theories which 
at least partially explain them. Its characteristic research approach is the use of intensive 
comparisons, which test a range of predicted contrasts or similarities across a small number of 



29 35 

theoretically sampled countries. The intensive approach compensates for the lack of degrees 
of freedom at the country level by making multiple comparisons and testing a range of 
hypotheses arising from the same theoretical starting point. The intensive approach is 
therefore dependent on detailed comparable data, and on a strong conceptual and theoretical 

Of the two main data sources used by the CATE WE project, only the Eurostat LFS provides a 
basis for extensive comparisons. It has standardised data on education and labour-market 
outcomes for all fifteen member states of the European Union, whereas the school-leaver 
survey data are only available for five countries. However a sample of fifteen countries is still 
a small sample to be used for extensive comparisons, especially when some countries have to 
be excluded from particular analyses for reasons of sample size or data availability, although 
the degrees of freedom can be increased by using data from a sequence of annual surveys. As 
a result, the EULFS is most effective in the analysis of transition systems when these are 
conceptualised in terms of a small number of types (such as OLM countries, ILM countries 
and southern European countries: see 3.5 below) or represented by time-varying variables 
such as the occupational distribution or level of educational participation. In the CATEWE 
project, analyses of the EULFS include: 

• mainly descriptive comparisons, for example of ‘dual statuses’ which combine education 
with work, or of trends in educational attainments or labour-market outcomes; 

• analyses which distinguish a small number of types of transition systems among EU 

• multi-level analyses which use country-level variables (usually including time-varying 
variables) along with individual variables, such as education, to predict such outcomes as 
unemployment or the type of employment of recent labour-market entrants. 

The second and third of these sometimes correspond to our notion of extensive comparisons. 
However, many of the analyses more closely resemble intensive comparisons, as they take 
advantage of the EULFS’ rich data across a range of labour-market outcomes in order to 
make the multiple comparisons more characteristic of the intensive approach. Further 
analyses use data from national labour force surveys in order to take advantage of the wider 
range of data and the scope for re-defining variables for more focused and detailed 



SLS data were available for only five countries, and for many analyses data were available 
only for a subset of these. The comparative approach in the SLS analyses has been intensive 
rather than extensive, focusing on multiple comparisons across a range of outcomes. An 
example is Smyth’s (2000a) analysis of the effects of varying levels of differentiation in 
education systems across a range of indicators of gender inequality - in the level of education, 
type of education, occupation, income, and so on. The SLS data for many transition processes 
and outcomes (the dependent variable) are incomplete or insufficiently comparable across 
countries (see 3.4 below). The school leavers' surveys support some important analyses but 
their potential for intensive comparisons, which require detailed, comparable data on a range 
of processes and outcomes, is more limited than we had hoped. On the other hand, the school 
leavers' surveys have proved unexpectedly valuable for what we have termed interpretive 
comparisons. The processes of constructing an integrated dataset, and of defining cross- 
nationally applicable measures of such concepts as the level of education or of educational 
attainment, raise issues of meaning and equivalence which are commonly overlooked in 
comparative research. And deeper investigations of concepts and institutions such as 
apprenticeship, youth programmes and upper-secondary vocational education draw attention 
to cross-national differences in their organisation and their role in the transition process. 

The project’s main comparative strategy is thus an intermediate one in terms of Figure 3.2. It 
uses common concepts to analyse country differences and (at least partly) to explain them, 
and it relies primarily on an intensive approach which tests multiple hypotheses from a given 
theoretical starting point on a small sample of countries. 


This section discusses the strengths and weaknesses of the two data sources used in the 
project. It has been pointed out in section 3.3 that the European Union Labour Force Survey 
(EULFS) can be used for extensive comparison while the national school leavers' (or 
transition) surveys are useful for intensive comparison. The data sources differ in other ways 
which are outlined in Figure 3.3 (see also Raffe, 2000). 

31 3 ? 

Figure 3.3: Comparison of data sources 


EU Labour Force Survey 
(uptoLFS 1997) 

National School Leavers' 


Nature of survey 

Cross-sectional (snapshot 
at one point in time) 

Longitudinal (flow out of 

Data structure 

Cross-sectional with only a 
limited retrospective 

More complete 
retrospective information 
on educational and labour 
force histories; some panel 
(follow-up) information 


At least annual 

Regularly, though not 
necessarily annual 

Country coverage 

All EU countries 

France, Ireland, the 
Netherlands, Scotland and 
Sweden 1 

Sample coverage 

All adults; allows for a 
comparison of young 
people and adults 

Young people 
experiencing the transition 


Constructed to a 
comparable framework but 
process inadequately 

Not designed to be 
comparable but useful 
comparative indicators can 
be constructed 

Form of data 

Aggregate only at EU 
level; micro-data for some 
individual countries 

Individual-level data 

Information on educational 

Level and type (general v. 

Level, type, field and 
grades (for some countries 
at least) 

Information on 
employment position 

Detailed: principal status 
and nature of job but lack 
of information on 
participation in youth 
programmes or earnings 

Detailed: principal status, 
nature of job, earnings, 
and participation in youth 

Information on social 

Not available 

Parental social class, 
parental education, 
immigration status (for 
some countries at least) 

Perhaps the key difference is that the EULFS gathers information on a cross-sectional basis 
collecting data on adults within households at a single point in time. The EULFS, therefore, 
has little direct information on the transition process itself, but researchers can compare 
different age groups within the labour force, or compare recent and earlier entrants to the 

' A number of other regional or sectoral studies were available elsewhere but were not suitable for inclusion in 
the project given its focus on national systems. 

labour market. In contrast, national school leavers' surveys explicitly take account of the 
transition process by examining the flow out of education into the labour market or further 
education/training. With school leavers’ survey data, one has the advantage of being capable 
of observing individual trajectories over a certain time-period after leaving the educational 
system. That is, individual labour force histories are directly observed as a sequence of labour 
force statuses and their associated features. Obviously, this is impossible from LFS data 
sources, as the necessary information is simply absent. Still, some insights into transition 
processes can be gained from generating aggregate career paths by comparing the 
distribution of labour market states between individuals having spent different amounts of 
time on the labour market. The basic difference between using SLS and LFS data for 
transition research thus lies in the fact that SLS allows us to represent individual transition 
processes, while LFS sources are restricted to analyses of aggregate (average) patterns only. It 
must be emphasised, however, that the two data sources yield complementary insights into the 
transition from school to the labour market. In the remainder of the section, we discuss these 
characteristics, advantages and shortcomings of the data sources in greater detail. 

3.4.1 Addressing transition processes from LFS data 

The EULFS appears as an attractive database to comparative research for a number of 
reasons. For each of the fifteen current member states, the LFS provides information based on 
large sample sizes, which allow for differentiated results on labour force activity and its 
determinants, and the surveys are explicitly administered according to a design which is 
geared towards producing comparable information across countries. In addition, the LFS 
surveys are repeated at least annually, so that they represent one of the few databases from 
which current processes of social change can effectively be studied. Last but not least, the 
LFS in general contains a wealth of information on current labour force activities, 
employment conditions (occupation, industry, hours, job tenure etc.) and socio-demographic 
characteristics of respondents. These advantages of the EULFS come at a price, however, at 
least in a study on school-to-work transitions (cf. the discussion in Cereq, 1997; Couppie and 
Mansuy, 1999). 

The main drawback of the LFS surveys, at least as currently released by Eurostat, is their 
purely cross-sectional nature at the level of individual respondents. That is, the EULFS by 
definition does not allow us to follow a dynamic research approach at the micro level which 

would investigate individual trajectories from education into the labour market. The necessary 
information is simply not present in the database as the same individuals cannot be followed 
over time. 2 Still, to a limited extent, retrospective information, on, for example, past 
employment status, is available from LFS sources. In fact, this information has been used in 
some analyses (Couppie and Mansuy, 2000a, 2000b), and some results from these will be 
presented below. On the other hand, it has to be recognised that the analytical value of the 
retrospective information provided is in itself quite limited: as retrospective measures on 
potential explanatory covariates are not available, causal analysis of labour market flows is 
greatly inhibited as it can only be conducted for those characteristics which are (reasonably 
assumed to be) stable over time. In addition, measurement concepts for labour force statuses 
are not identical at both time points, so that definitional problems might plague any such 
analysis. Using EULFS data for transition research for most purposes thus has to represent a 
conscious decision to restrict one’s analytical potential to what is available from cross- 
sectional data while being able to cover all EU countries in the research. Consequently, the 
LFS analyses performed within the CATEWE project can be read as an attempt to extract as 
much information on youth labour market integration in Europe as possible from this cross- 
sectional database. 

The key to our analyses is to realise that, although it is impossible to observe individual 
trajectories between education and work in the LFS data, cross-national similarities and 
differences in macro-level patterns of youth labour market integration can readily be observed 
from the database. In fact, as information is collected annually, traditional cohort analyses can 
be performed if information from subsequent surveys is linked accordingly. Muller and 
Wolbers (1999) applied this technique to address the scope of educational expansion in 
Europe over the past decades: by combining educational distributions for the same birth 
cohorts over the historical observation period currently available from the EULFS, they have 
been able to analytically separate life-cycle patterns of educational attainment from cohort 
effects on the level of educational attainment. For the purpose of country comparison, this 
generates a valid picture of educational attainment processes at the macro level, even without 
the availability of longitudinal data at the individual level and without the imposition of any 
additional assumptions on the data. 

2 While there are some national LFS studies which actually do, and others which at least in principle would 
permit us to follow this approach, the harmonised EULFS does not so far allow the identification of members of 
existing rotating samples across annual survey waves. 

A n 




This approach can, in principle, also be applied to the analysis of labour market outcomes for 
those entering the labour force. On the other hand, analyses based on genuine birth cohorts 
are unlikely to yield adequate (comparative) representations of labour market entry processes: 
within any single country, leavers from different educational backgrounds exit the education 
and training system at different ages. That is, any straightforward differentiation of youth 
labour market outcomes by age tends to misrepresent the situation of interest as young people 
of a given age might actually be in very different career stages. A university graduate might 
have just begun working in her first job at age 27, while somebody who left school after 
attaining his compulsory education certificate has already been working for 10 years. The 
issue becomes even more problematic in comparative research, as the precise biographical 
timing of these transitions varies according to the particular institutional structure of national 
education and training systems. Moreover, differences in national levels of educational 
attainment can generate misleading country differences in macro-level indicators of labour 
market outcomes, totally unrelated to any differences in actual integration processes. For 
these reasons, most of the project’s analyses of labour market outcomes are based on labour 
force entry cohorts rather than birth cohorts. That is, the incidence of the transition period has 
been defined relative to the biographical time point of leaving the education and training 
system rather than sheer biological age. In fact, this defines our perspective on the transition 
period as one on labour market outcomes in the early career stage, after having completed 
initial education and training. In most analyses, we chose typical graduation ages for different 
types and levels of education as published in the OECD's Education at a Glance series (see, 
for example, OECD, 1997) as an approximation to the biographical time point of completing 
a particular type of education in the various European countries, and calculate a measure of 
potential labour force experience on that basis. For the purposes of the project, we then focus 
on labour market outcomes among those in their initial years on the labour market, that is, up 
to five, or in some analyses up to ten or even fifteen years after having obtained their highest 
educational qualification. In addition, most analyses assume relative stability in the structure 
of transition processes in the short-run so as to enable the use of synthetic cohort approaches 
in the statistical analyses. By doing so, we are able to extract the macro level properties of 
transition patterns for all current EU member states from LFS databases. 3 

3 Two papers actually used national LFS microdata for some in-depth analyses of particular questions in a more 
restricted sample of countries (Brauns et al., 1999; lannelli and Soro Bonmati, 2000), but the methodological 
remarks made here apply to these analyses as well. 

Certainly, this still represents a serious limitation with respect to an adequate description of 
occurring transition processes. It is most discomforting not to be able to describe individual 
level transition processes because the full extent of individual heterogeneity in transition 
processes between education and work cannot possibly be uncovered from LFS data. On the 
other hand, what still can be observed from this database under the chosen set-up, is the 
average labour market outcomes of young people in their early career stages in all European 
countries and conditional on education, time since leaving education and training systems, 
and certain socio-demographic factors. Exploring the incidence of cross-national differences 
in such average transition outcomes is actually a major task in understanding the outcomes of 
different institutional arrangements regulating school-to-work transitions in Europe which has 
not been done in a similarly encompassing fashion so far. 

Of course, more detailed longitudinal data do allow for more detailed studies of transition 
processes. In particular, using the LFS implies the restrictive assumption that (a) leaving the 
education and training system, (b) achieving the highest qualification and (c) entering the 
labour market all happen at the same time; in other words, that only the most recent of 
potentially multiple entries is of substantive importance. As we argue in 3.7 below, this need 
not be the case: moreover, the relationship between these three events is variable across 
countries. Nevertheless analyses based on this simplified assumption can provide useful 
insights into cross-national differences, which are consistent with project research based on 
school leaver survey data as well as previous research, as the work summarised in section 3.5 
below demonstrates. 

3.4.2 Analysing transition data using national transition surveys 

One of the key objectives of the CATEWE project was to use existing national transition 
surveys to explore early transition processes among young people in a range of countries. In 
most of the countries concerned (France, Ireland, the Netherlands and Scotland) these surveys 
were surveys of 'school-leavers', that is, those exiting secondary education at a particular 
point in time, although the precise definition of a 'leaver' can vary across countries. 4 However, 

4 In the Scottish context, for example, 'leaver' refers to those leaving the general secondary school system with 
further education regarded as a destination. Countries also vary in their approach to apprentices who are 
differentially treated as 'leavers' and 'labour market entrants'. 

in the Swedish case, data related to a series of cohort surveys of young people leaving 
education at various stages. National transition surveys have a number of advantages in 
exploring educational outcomes and early labour market experiences among young people 
(see Figure 3.3 and Raffe, 2000). Firstly, they tend to collect detailed information on 
educational background, incorporating dimensions which are considered important in the 
particular institutional context. The sensitivity of transition surveys to the national context is 
an advantage in providing a more complete view of (national variations in) the transition 
process. Secondly, they allow us to directly relate young people's educational background to 
their experiences in labour market integration at an individual level. Thirdly, the fact that they 
are (for the most part) leavers' surveys means that we are looking at young people, most of 
whom entered the labour market at the same time and therefore searched for work and started 
their careers under the same conditions. Fourthly, such surveys tend to provide rich data on a 
range of transition 'outcomes' among young people, covering not just principal activity but 
different dimensions of job quality. 

National transition surveys have, however, been largely under-exploited in cross-national 
analyses. One of the main advances made by the CATEWE project has been, therefore, to use 
these national transition surveys to construct an integrated database with information on key 
aspects of young people's educational and labour market experiences. In fact, a total of three 
databases were constructed as part of the project: 

1. A current database. This was based on the most recent year for which school leavers' 

surveys were available. This database covers: 

• France: young people who left general or vocational full-time secondary education 
(including apprenticeships undertaken as part of initial education, but excluding 
General Baccalaureat and agricultural courses) in 1993-4 and who did not continue in 
full-time education; 

• Ireland: young people who left secondary school in 1995-6, surveyed in autumn 1997. 
This includes those who left Junior or Leaving Certificate and Post-Leaving 
Certificate courses. Other post-secondary vocational courses count as destinations, 
along with apprenticeships, training schemes and third-level education. 

• The Netherlands: young people who left secondary education (including MBO) in 
1995-6 and did not enter another form of secondary education. They were surveyed in 
autumn 1997. In this survey, apprenticeships count as destinations. 





• Scotland: young people who left general secondary school in 1993-4, surveyed in 
spring 1995. Courses in further education colleges, apprenticeships, training schemes 
and higher education courses count as destinations; 

• Sweden: young people who completed lower-secondary (compulsory) education in 
1993 and were surveyed in spring 1997. 

This database has been used to examine cross-national differences at a very early point in 
the transition process, one to one and a half years after leaving school. 

2. A time-series database. Since no comparable full leavers' survey was available for earlier 
years for France, it was based on three countries: Ireland (1980-1997; 5 time-points), 
Scotland (1979-1995; 5 time-points), Netherlands (1989-1997; 3 time-points). This 
database has been used in order to explore changes in institutional and labour market 
contexts over time. 

3. A longitudinal database. Since the current and time-series databases relate to a very early 
point in young people's labour market career, this has the advantage of allowing the 
analysis of longer term transition patterns. Unfortunately, due to lack of data availability, 
the construction of a longitudinal database was only possible for Ireland and France, and, 
for a much more limited set of variables, for France and Sweden. 

Constructing the integrated databases was not just a means of conducting substantive analyses 
on school to work transitions in a number of countries but the process in itself yielded 
valuable insights into our understanding of institutional variation in education, training and 
labour market systems. The procedure involved an iterative process, involving the 
identification of common variables within the national datasets, the development of a 
'mapping' from the original, often highly diverse, country-specific variables to a common 
variable specification and the testing of these new comparative variables (for further details, 
see CATEWE, 1999). This approach yields a number of advantages over the analysis of 
transition surveys at a national level (see Brannen, Smyth, 2000; CATEWE, 1999). It allows 
us to directly test cross-national differences in transition patterns, controlling for a range of 
other factors. Thus, we examine whether, for example, educational relativities in 
unemployment differ between Ireland and the Netherlands, all else being equal. More 
importantly, the construction of comparable variables for countries with very different 
institutional contexts requires a rigorous clarification of the different dimensions of education 
and transition outcomes explored in the analyses. Thus, the work served to challenge our pre- 



A A 

existing assumptions about the nature of cross-national variation and contributed to the 
development of new classificatory schema for analysing different dimensions of the transition 
process. In particular, the construction of variables specifically for the purposes of the project 
has meant that we can directly reflect the central research questions we seek to address rather 
than using pre-existing (and often inappropriate) classification systems. It has also helped to 
develop a set of multi-dimensional indicators which better reflect the specificities of the 
different institutional systems. 

However, national transition surveys and any integrated database drawing on these surveys do 
have some limitations. They cannot allow us to compare the experiences of young people 
with older age-groups or with those who entered the labour market at a much earlier point in 
time, analyses that can usefully be carried out using the Labour Force Surveys. Thus, it can be 
difficult to distinguish whether cross-national differences in the employment experiences of 
young people reflect differences in the labour market structures as a whole or in the relative 
position of young workers vis-a-vis the adult population. Furthermore, differences between 
the national surveys in design and content result in difficulties in comparability (see 
CATEWE, 1999). These difficulties are as much conceptual as technical since, in the context 
of significant institutional differences between transition systems, it is impossible to identify a 
single transition event that has equivalent significance in each system and which can provide 
the basis for comparison. Finally, such surveys are available only for a (limited) number of 
countries: France, Ireland, the Netherlands, Scotland and Sweden. Thus, the database does not 
include any 'dual system' country or any country from Southern Europe, groups of countries 
which have been found to have distinctive profiles in terms of transition processes (see Muller 
et al., 1999). The set of countries studied includes one country often grouped with the dual 
system countries in terms of the predominance of occupational labour market arrangements, 
the Netherlands; the remainder of the countries, however, come from the group of North- 
Western European countries usually characterised as 'ILM' (internal labour market) countries 
(see, for example, Gangl, 1999). This distinction has also been characterised as the difference 
between systems with an underlying 'employment logic' and those with an 'education logic' 
(Iannelli and Raffe, 2000). However, the rich data from the national surveys allow us to 
explore potential heterogeneity among transition systems that may resemble each other in 
other respects. 



39 45 

In the following sections of the chapter, we highlight the main findings of the analyses using 
Labour Force Survey and national school leavers' data. 


CATEWE aimed to deliver a genuine European perspective on transitions from education to 
working life. This promise is not easily fulfilled as adequate longitudinal data from which to 
study labour market integration processes at the individual level are available for a limited 
subset of European countries only. Therefore, CATEWE has attempted to complement its 
analyses based on longitudinal data by analyses drawing on the European Union Labour 
Force Survey (EULFS). These analyses are first intended to provide a broader picture on 
patterns of labour market entry across EU countries, including those where longitudinal 
microdata was unavailable to the project. Analyses of EULFS data provide a unique 
opportunity to situate results from the analysis of School Leaver Surveys within an even 
broader European context. But apart from this purpose, the project also attempted to make use 
of the genuine potential of the EULFS data base for transition research. To do so based on 
cross-sectional LFS data is certainly less obvious than from a truly longitudinal database, but 
we believe that the EULFS sources have some inherent qualities of their own in that respect, 
which can be fruitfully exploited by proper statistical analysis. 

Within the project, ten substantive working papers suitable for later scientific publication 
have been produced based on LFS data (see Table 3.1 below). Given the available 
information in LFS sources, most of the papers centre around the education-employment 
linkage in European countries. Individual papers explore, for example, cross-national 
similarities and differences in the educational background of young people entering the labour 
market, as well as the nature and scope of educational expansion over the past decades 
(Muller and Wolbers, 1999), or the provision of dual forms of vocational training in European 
countries and their evolution over the last decade (Wolbers, 2000). On the labour market side, 
there are papers aimed at a broad descriptive overview of labour market outcomes for recent 
entrants into the labour force (e.g. Couppie and Mansuy, 2000a), which in part also provide 
country classifications in terms of relatively similar aggregate transition patterns (Couppie 
and Mansuy, 2000b; Gangl, 2000a). In addition, there is a set of more analytical papers using 
advanced multivariate statistical techniques for causal analyses of unemployment risks and 





employment outcomes among labour market entrants in different European countries (van der 
Velden and Wolbers, 2000; Gangl, 2000b, 2000c). And finally, there are two papers which 
use more detailed national LFS microdata in order to understand the peculiarities of transition 
outcomes in Southern Europe (Iannelli and Soro Bonmati, 2000) and the effects of education 
on unemployment processes (Brauns et al., 1999). 


Hildegard Brauns, Markus 
Gangl, and Stefani Scherer 

Thomas Couppie and 
Michele Mansuy (2000a) 

Thomas Couppie and 
Michele Mansuy (2000b) 

Markus Gangl (2000a) 

Markus Gangl (2000b) 

Markus Gangl (2000c) 

Cristina Iannelli and 
Asuncion Soro Bonmati 
( 2000 ) 

Walter Muller and 
Maarten Wolbers (1999) 

Rolf van der Velden and 
Maarten Wolbers (2000) 

Maarten Wolbers (2000) 

Table 3.1 

Overview of LFS Working Papers 


Education and unemployment: Patterns 
of labour market entry in France, the 
United Kingdom, and West Germany. 

The Position of New Entrants on 
European Labour Markets. 

New Entrants and experienced workers 
on European Labour Markets. 

European Perspectives on Labour 
Market Entry: A Matter of Institutional 
Linkages between Training Systems and 
Labour Markets? 

Education and Labour Market Entry 
across Europe: the Impact of 
Institutional Arrangements in Training 
Systems and Labour Markets. 

Changing Labour Markets and Early 
Career Outcomes: Labour Market Entry 
in Europe over the Past Decade. 

The Transition from School to Work in 
Southern Europe: The Cases of Italy 
and Spain 

Educational attainment of young people 
in the European Union: cross-country 
variation of trends over time. 

The integration of young people into the 
labour market within the European 
Union: the role of institutional settings. 

Learning and working: Double statuses 
in youth transitions within the European 

Main Topics 

Role of education for avoiding extensive 
periods of initial job search and 
subsequent job instability in three 
European countries 

Overview of labour market outcomes 
among recent entrants to European 
labour markets 

Cross-national similarities and 
differences in various aspects of 
transition patterns, e.g. dual status 
situations, unemployment and 
employment outcomes 

Cross-national similarities and 
differences in the relations between 
labour force experience, qualifications 
and unemployment and employment 

Cross-national similarities and 
differences in the role of education for 
unemployment and employment 

Effects of macroeconomic and macro- 
structural trends on transition outcomes 
in Europe 

Comparison of the patterns of transition 
from education to the labour market in 
Spain and Italy 

Cross-national similarities and 
differences in educational backgrounds 
of market entrants in Europe; similarities 
and differences in the nature of 
educational expansion 

Effects of institutional context factors on 
unemployment and employment 

Cross-national similarities and 
differences in the incidence of combined 
training and work activities in Europe 


The following represents an attempt to provide a concise summary of the main results from 
our analyses. Rather than summarising individual papers, this review will be organised along 
the substantive themes covered, which have been touched upon in one or more of these 
analyses. More specifically, project results are reviewed for the issues of (a) educational 
achievement and the nature of qualifications among school leavers in Europe, (b) cross- 
national similarity and difference among European countries in terms of transition patterns 
from education to work, (c) the incidence of unemployment among market entrants and the 
role of educational, institutional, and other contextual determinants, and (d) the nature of 
employment outcomes in early career stage and their determinants across European countries. 

3.5.1 Educational achievement and the nature of qualifications across Europe 

In a certain sense, education is the key resource available to individuals to influence their 
labour market fortunes. Education and training represents an individual investment in 
qualifications which are afterwards rewarded on the labour market. This general relation 
already bears on the nature of education-to-work transitions as different types of training 
might be more or less effective in generating smooth patterns of labour market entry. To the 
extent that training systems vary between countries, the different resources institutionally 
provided to young people can be expected to lead to substantially different transition patterns 
in different European societies. And indeed, as amply shown in project research (notably 
Muller and Wolbers, 1999; Wolbers, 2000), the qualificational background of labour market 
entrants is strikingly different across European Union countries. These differences relate to 
both the level of educational attainment as well as the nature of qualifications obtained. More 
specifically, three broad country patterns seem to emerge from our data, which closely reflect 
the underlying institutional structures in education and training systems (Muller and Wolbers, 


As a first type, there are those Continental countries operating extensive vocational training 
systems at the upper secondary level, like Austria, Denmark, Germany, and the Netherlands, 
but also the other Nordic countries Sweden and Finland. In all of these countries, the 
proportion of young people not progressing beyond compulsory education levels is very low, 
typically well below 15 per cent of a birth cohort. At the same time, a significant proportion 
of young people, typically 25 per cent of a cohort and more (with the exception of Austria), 
obtain tertiary level qualifications. But the most distinctive feature is the fact that almost 


A r\ 

everybody who left the educational system from the upper secondary level will have obtained 
occupationally-specific qualifications. Of course, these will mainly have been acquired in the 
context of dual system arrangements in Austria, Germany, and, to a lesser extent, in Denmark, 
while the dominant vocational training route will be school-based training in the Netherlands, 
Sweden and Finland. Compared to these countries, the aggregate pattern of educational 
attainment is somewhat different in the remaining Northern European countries, though. 
Broadly speaking, there is little difference between Northern European countries in terms of 
tertiary level graduation rates. The UK, and more so Ireland, France, and Belgium differ from 
the former set of countries mainly in the fact that fairly large proportions of upper secondary 
level leavers enter the market with general rather than vocational qualifications. In addition, 
the progression beyond compulsory education is significantly lower in these countries than in 
the other Northern European countries. In fact, Southern Europe constitutes a third empirical 
pattern, mostly distinguished from countries like the UK or France by the lower level of 
educational attainment (except Spain), rather than any difference in the vocational-general 
mix at the upper secondary level. If anything, then Southern European education and training 
systems provide even less vocationally-specific training than is the case in most Northern 
European countries. As an illustration to these distinctions, Figure 3.4 below depicts the life- 
cycle pattern of educational attainment in six exemplary European countries (Muller and 
Wolbers, 1999). 



4 $ 

Figure 3.4 

Educational Attainment by Age, Selected European Countries 




United Kingdom 

Source: Muller and Wolbers, 1999. 

In part, these country differences are rapidly changing. As the analyses of Muller and 
Wolbers (1999) show, the nature and pace of educational expansion has varied significantly 
between European countries over the past two decades. Those countries, for example, which 
previously had the highest proportions of individuals with only compulsory education were 

also the most successful in reducing these proportions recently, while the Nordic countries or 
Germany and Austria have been much less able to reduce these figures below the levels 
already achieved a generation ago. In much the same vein, catching-up processes also 
occurred at the higher levels of education. Recent educational expansion of tertiary education 
has been occurring fastest in Southern European countries, notably in Spain and Portugal. 
Similarly large expansions took place in Ireland, France and the UK, while the respective 
trends have been much weaker in Austria, Germany and the Nordic countries. In sum, there 
are considerable trends under way towards converging educational levels among young 
people in different European countries. 

While the levels of educational achievement may actually converge somewhat across 
European countries, it is much less likely that the more specific nature of initial qualifications 
provided will actually converge quickly. As has been indicated above, a crucial distinction 
between European countries is the extent to which education and training systems already 
provide occupationally-specific training (mostly at the upper secondary level). Most 
distinctive to systems providing occupationally-specific training are large-scale dual system 
arrangements as operated in Austria, Germany, but also school-based vocational training in 
the Netherlands. One might actually argue that these systems are merely an institutionally 
different solution for providing adequate training to young people. In cases where a dual 
system exists, training provision is more regulated and integrated more closely into the 
education system, while in countries lacking such arrangements, the respective training is 
provided by companies under their own auspices. 

Results from the project cast some doubts on such optimistic perspectives, however. 
Unsurprisingly, an analysis by Wolbers (2000; see Table 3.2 below) clearly shows that 
participation in dual system training occurs most often in those countries operating large-scale 
dual systems. To a large extent, participants come from compulsory education backgrounds, 
that is, they participate in dual system training as a means to progress beyond the lowest level 
of education. Consistent with the above notion, Wolbers then also establishes a slightly higher 
tendency for Northern Europeans outside the core occupationalised systems to combine 
regular employment with further education. But apparently, this training occurs mostly among 
tertiary level graduates rather than the lowest qualified — effectively, it is thus very unlikely 
that dual system training foregone is made up by company training for the least qualified later 
on. That is, it is typically not those leaving from compulsory levels of education who receive 


K -f 


subsequent company investments, but rather those individuals who already bring a high level 
of qualifications to the work place. To do full justice to Wolbers' results, one should also note 
that precisely those Northern European countries lacking large scale vocational training 
arrangements have been those which significantly extended the provision of training which 
combines learning and working. Neither in the traditional occupationalised systems nor in 
Southern Europe did the proportion of young people receiving such training change 
substantially over the last decade. In Southern Europe, in particular, the likelihood of 
receiving occupationally relevant training after leaving the education and training system is 
very low. 

Table 3.2: 

The Structure of Combined Work-Training Activities, by Institutional Contexts 


Macro-institutional Context 


Wolbers, 2000 

Dual System Countries (incl. NL) 

highest probability of dual system training, stable over 
time; less strong gender-typing of dual system training 

Southern Europe 

lowest probability of dual system training, working 
students and further education among employed 

other European countries 

relatively low probability of dual system training, but 
increasing over time; highest probability of further 
education among employed 

3.5.2 Cross-national similarity and difference in transition outcomes: a broad 
perspective on labour market entry patterns in European countries 

Conditional on leaving the education and training system, what are the similarities and 
differences between European countries in terms of labour market outcomes which occur on 
entering the labour force? And if differences occur, are there some countries which are 
relatively similar to each other in terms of the observed outcome patterns while others differ? 
Which contours does a European map of transition experiences show? In fact, these questions 
can be answered along numerous dimensions, each emphasising a particular aspect of labour 
market and employment outcomes. Within the project, most research has focused on 
unemployment on the one hand and job features like occupation, industry, and type of 
contract on the other, although alternative measures have also been considered to some extent 
(Couppie and Mansuy, 2000a, 2000b; Gangl, 2000a). In addition, these analyses attempted to 
arrive at a descriptive account of major similarities and differences in transition patterns 




between European countries, often yielding empirical classifications of countries according to 
observed similar transition patterns. Within the project, several approaches have been 
followed, focusing either more on cross-national variation in outcome distributions like 
proportions of dual system training, youth unemployment rates or average occupational status 
outcomes (Couppie and Mansuy, 2000b), or more on cross-national variation in the relations 
between qualifications, labour force experience and employment outcomes (Couppie and 
Mansuy, 2000a; Gangl, 2000a). In doing so, attention was given to both comparative 
perspectives on the features of national youth labour markets and the relationships between 
youth and adult labour markets. 

In fact, there are some broad cross-national similarities between European countries in terms 
of labour market experiences among recent entrants to the market, notably as compared to 
those of more experienced workers (Couppie and Mansuy, 2000a, 2000b; Gangl, 2000a). 
Typically, unemployment rates are higher at the early career stages as people have to look for 
a first job or have been able to secure only fairly uncertain jobs in the beginning (Couppie and 
Mansuy, 2000a, 2000b; Gangl, 2000a). Similarly, those entering the market are 
disproportionately allocated to low-skilled service-sector jobs (Couppie and Mansuy, 2000a, 
2000b; Gangl, 2000a), often under fairly precarious contract conditions, as signified, for 
example, by the higher incidence of fixed-term contracts among market entrants (Couppie and 
Mansuy, 2000a). Moreover, transitions between labour market statuses of employment, 
unemployment, and inactivity occur much more often among market entrants compared to 
more experienced workers (Couppie and Mansuy, 2000a). 





Figure 3.5 

Unemployment Rates and Labour Force Experience 

'Austria *0— Denmark 

■Germany -^r- 1 Netherlands 

—O— Belgium 



—^—United Kingdom 

-O-Greece ""O— Italy | 
^^"Portugal Spain | 

Notes: Leavers from ISCED level 3; lines represent results from logarithmic smoothing. 
Source: Gangl, 2000a. 

On the other hand, it is important to recognise that national transition patterns are far from 
identical, even if some aspects are common to most, if not all, of them. Indeed, countries 
differ markedly with respect to some core aspects of youth transition experiences. There are 
some countries where unemployment risks among market entrants are markedly more 
pronounced than those for more experienced workers. The Southern European countries, but 
also France, are examples for these (cf. Figure 3.5 above as an illustration for the group of 
leavers from upper secondary (ISCED level 3) education). But there are also other countries 
where this relationship is extremely weak, so that unemployment rates among market entrants 
closely parallel those among more experienced workers. Austria, Denmark, Germany, and the 
Netherlands would be examples of the latter group of countries (Couppie and Mansuy, 2000a, 
2000b; Gangl, 2000a). In much the same way, the degree of disproportionate allocation to 
lower-level employment or to the service sector varies between European countries: while 
many young people enter the labour market at particularly low job levels and then progress 
over their initial years in the labour market in terms of occupational status or similar measures 
of job characteristics, this tendency is significantly weaker in Germany or Austria, for 
example (Couppie and Mansuy, 2000a; Gangl, 2000a). Also, it seems relatively common 



K. A 

among European countries, that a considerable proportion of new entrants to the market enter 
non-standard forms of employment, which are then increasingly left over the initial years in 
the labour force. Typically, some 20 per cent of an entry cohort held temporary contracts in 
their first year on the market, although the Nordic countries, and even more so Spain exhibit 
markedly higher figures, with estimates ranging even up to 80 per cent in the Spanish case 
(Couppie and Mansuy, 2000a). In many respects, similar patterns are also evident in the case 
of part-time employment. While part-time employment in general is much less specific to the 
early career stage, Couppie and Mansuy (2000a) show that the incidence of involuntary part- 
time contracts clearly declines with increasing labour force experience in almost all European 
countries. But again, countries differ remarkably in the extent to which young people have to 
accept involuntary part-time employment: notably in Belgium, France, Sweden and Finland 
the respective proportions amounted to well above 10 per cent among new entrants to the 
labour market. 

These findings of important heterogeneity among the countries also extend to particular types 
of labour market mobility. While it is true for many countries that young people are faster to 
leave unemployment, there are important exceptions to this rule. In neither Italy nor Greece 
does the likelihood of leaving unemployment vary by experience, and the same holds for the 
UK; at the same time, the transition rates from unemployment to employment in the UK are 
about twice those for Greece and Italy (Couppie and Mansuy, 2000a; cf. Figures 3.6 and 3.7 
below). Similar observations can be made with respect to the probability of losing 
employment and subsequently entering unemployment, where most countries exhibit a 
modestly negative relationship with increasing work experience. Spain and France, in 
particular, experience excessively high inflow rates among market entrants, however 
(Couppie and Mansuy, 2000a; cf. Figure 3.6 below). Behind all these descriptive findings, the 
main substantive result is to realise that European countries differ much less in terms of 
labour market outcomes among experienced workers than they do in terms of outcomes 
among market entrants. To the extent labour market entrants' fortunes differ across European 
countries, this reflects cross-national variation in the relative competitiveness of those leaving 
the education and training systems, that is, the extent to which market entrants achieve similar 
outcomes as experienced workers along a number of dimensions. Variation in this 
relationship is at the core of empirically distinguishable 'transition systems' among European 

Figure 3.6: Job exit rates among individuals employed in the previous year, by labour force experience 






Sources: Couppie and Mansuy, 2000a 

The case for institutional explanations of market entrants' competitiveness in different 
countries could be strengthened if one were able to show that countries exhibiting similar 
institutional frameworks in training systems and labour markets are actually relatively similar 
in such overall transition patterns. In our analyses, we have attempted to demonstrate this by 
exploring the nature of cross-national similarities and differences on a number of labour 
market dimensions from cluster analyses (Couppie and Mansuy, 2000b; Gangl, 2000a). And 
although the technique is exploratory in nature, it does show some intriguing profiles of 
European differences in transition patterns. Our different analyses reliably singled out two 
polar transition patterns deviating clearly from the rest of Europe: the occupationalised 
systems of Austria, Denmark, the Netherlands, and Germany, on the one hand, and the 
Southern European countries, including Italy, Greece, and Portugal at least, on the other. 

The main features of the first ideal-typical pattern was that, by and large, market entrants 
achieve only slightly less favourable market outcomes than more experienced workers, in 
terms of both unemployment and job characteristics. In contrast, in countries like the UK, 
France and Ireland, market entrants are significantly disadvantaged compared to more 
experienced workers on both dimensions. The Southern countries, in turn, deviate from this 
pattern by even more marked disadvantages to market entrants in terms of unemployment 
risks, but even more so in the low level of mobility between labour market statuses, once 
initial employment has been secured. Labour markets in Northern European countries exhibit 
much larger mobility rates between employers and between employment and unemployment 
than is the case in the typical Southern European experience. In fact, this criterion yields a 
major reason for considering youth experiences in, for example, Spain as relatively similar to 
France rather than to Portugal. 

Compared to the polar cases, further divisions among the remaining European countries 
emerged less clearly. It is clear that the remaining countries comprising France, the UK, 
Belgium, Ireland, Spain, but possibly also Sweden and Finland, form a much less 
homogeneous set of transition profiles than those described earlier. It is certainly possible to 
draw finer distinctions among transition patterns in these countries, but pur research based on 
LFS data has so far not given definite results. Depending very much on the particular 
indicators considered, sometimes Britain and Ireland could be distinguished from a particular 
French pattern, but in other analyses Ireland became included among the Southern European 


51 58 

countries. Clearly, this uncertainty in the results reflects, to a large part, the exploratory 
nature of the particular methodology applied and differences between the analyses in terms of 
the precise indicators utilised. But in fact, as some fairly clear-cut broad types of different 
national transition patterns seem to characterise the overall European experience, there 
appears some scope for institutionally-based explanations of these patterns. It is to these to 
which we now turn. 

3.5.3 Unemployment risks: education, institutions, and socio-economic context 

Among the various aspects of school-to-work transitions of potential interest, the project’s 
more specific and most sophisticated analyses have focused on two core labour market 
outcomes, namely unemployment risks and employment outcomes among labour market 
entrants. In the respective analyses, we attempted to explain these transition outcomes by 
adequately accounting for both the role of individual resources and characteristics and the 
impact of particular institutional contexts and other macrostructural and socio-economic 
context conditions (van der Velden and Wolbers, 2000; Gangl, 2000b, 2000c). In order to 
properly accomplish this task, the three relevant papers applied several variants of multilevel 
analysis as a methodological innovation in comparative empirical research. While this section 
will summarise our results with respect to determinants of unemployment risks, the 
subsequent section will discuss the determinants of specific employment outcomes in greater 


Among the many factors which could potentially be linked to the incidence of unemployment, 
our results mainly concern three types of determinants: individual education and training 
(Gangl, 2000b; Brauns et al., 1999), institutional features of both education and training 
systems and labour markets (van der Velden and Wolbers, 2000; Gangl, 2000b), and socio- 
economic context conditions (Gangl, 2000c). Unsurprisingly, each factor turns out to have 
important consequences for the extent of labour market integration problems among recent 
entrants. Education and training, for example, is the primary individual resource for avoiding 
unemployment at entering the labour market. In general and controlling for other factors, the 
higher the individual level of education attained, the lower the risk of unemployment 
incidence in the early career stage. At the same time, vocational training, notably if obtained 



in the context of dual training arrangements, also contributes to lower unemployment risks 
(van der Velden and Wolbers, 2000; Gangl, 2000b). Figure 3.8 below represents these 
relationships graphically, based on a multilevel model which controls for other individual 
factors, as well as institutional and economic context factors. 

However, the interesting result is that this relationship does not hold in all European 
countries; rather, there is systematic institutional variation as to whether educational 
credentials serve to lower the individual risk of unemployment. In fact, there is little variation 
among, broadly speaking, Northern European countries, except for the better performance of 
those who achieve school-based vocational training in occupationalised labour market 
contexts like Germany and the Netherlands. That is, apprenticeships and similar types of dual 
system training lead to lower unemployment rates than those of upper secondary general 
tracks in both occupationalised and less occupationally structured systems. The difference 
between these two types of transition systems lies in the fact that leavers from school-based 
vocational training face lower unemployment rates in more occupationalised contexts. This 
finding seems to support the reasoning that appropriate vocational specialisation is important 
to integrate young people into the labour force in markets exhibiting strong occupational 
boundaries (irrespective of whether the qualification is obtained from school-based or dual 
forms of training), while in less tightly structured systems it is more the actual training 
contract with a particular employer (as an apprentice or otherwise) which reduces subsequent 
unemployment risks. 

But the main institutional divergence occurs in Southern European countries, where the level 
and type of education hardly affects unemployment risks at all. That is, while low qualified 
school leavers do not face particularly different unemployment risks compared to their 
Northern European counterparts, unemployment rates among leavers from upper secondary 
education and even among university graduates are at similar levels to, rather than 
substantially lower than, those among the least qualified. In contrast to Northern Europe, 
unemployment in the early career stages is a particular problem of the highly qualified in the 
South rather than among the. least qualified. In fact, the nature of unemployment itself is thus 
likely to be very different between Northern and Southern Europe. 





Figure 3.8 

Unemployment among Market Entrants: Effects of Education and Institutional Context 

Lower secondary Apprenticeships Upper secondary Upper secondary Lower tertiary University degree 

vocational general 

Notes: Predicted probabilities at mean individual covariates and macrolevel context conditions, based on 

multilevel regression estimation. 

Source: Gangl, 2000b 

In order to deepen our understanding about the role of education in actual unemployment 
processes underlying the above results, we have conducted a more sophisticated analysis of 
educational effects on unemployment processes based on LFS microdata for France, the 
United Kingdom, and Germany (Brauns et al., 1999). In that analysis, we estimated a two- 
stage model of the labour market entry process, distinguishing between unemployment risks 
due to prolonged initial job search and unemployment risks related to the instability of initial 
employment found. From this analysis, it appeared that educational resources have 
reinforcing effects on both stages, that is, those qualifications which provide relatively 
smooth access to first jobs also typically provide access to more stable employment. This 
applies in particular to apprenticeships, which are found to provide not only almost immediate 
access to employment (e.g. by continued employment in the training firm), but also relatively 
secure first jobs. 

Apart from this different role of education in labour market allocation processes, the 
institutional structure of education and training systems actually exerts a crucial influence in 





itself as it determines the qualificational resources available to labour market entrants to a 
large extent. Those countries operating large-scale dual systems of training provision 
experience significantly lower unemployment rates in the transition period because a large 
proportion of those leaving the education and training system have acquired a qualification 
which implies direct access to subsequent employment (van der Velden and Wolbers, 2000; 
Gangl, 2000b). According to estimates from Gangl (2000b), this effect alone amounts to 
lower aggregate unemployment rates among labour market entrants in dual system countries 
(including the Netherlands) by 5 percentage points as compared to the other Northern 
European countries. In addition, there is also a composition effect of educational levels on 
unemployment rates, which further disadvantages the Southern European countries as 
compared to Northern Europe: the higher the level of education among market entrants, the 
lower a country's unemployment rate among this group. Beyond these institutional effects of 
educational systems and broad labour market contexts, there is little evidence for other 
relevant institutional factors. Van der Velden and Wolbers (2000) tested for effects of a 
number of institutional features of labour markets, including wage bargaining structures, 
union density, the extent of youth activation and training measures among others, but none of 
these receives clear empirical support. Only in the case of the strictness of employment 
protection legislation, they find evidence for a small positive effect on unemployment. That 
is, the better protected the core work force, the more difficult it is for youth to successfully 
compete in securing employment. On the other hand, this effect did not receive clear support 
as soon as the structure of training systems was simultaneously controlled for. 

Table 3.3 

Institutional Effects On Unemployment Rates among Market_ Entrants^ 


Institutional Variable 


van der Velden / 
Wolbers, 2000 

Centralised Wage Bargaining 
Union Density 

negative, not significant 
no effect 

Employment Protection 

positive, not significant 

Vocational Specificity / Educ. 

no effect 

Dual System 


Tracking / Second. Educ. System 

no effect 

Gangl, 2000b 

Apprenticeship Systems 


Occupationalized Markets 

negative (for vocational qualifications) 

Southern Europe 

positive (for better qualified) 

Gangl, 2000c 

Interaction of macroeconomic trends and 
three macroinstitutional contexts 

cyclical effects less pronounced in Southern 





At the same time, young people's labour market fortunes are not isolated from the evolution 
of the labour market in general. It is not only individual qualifications and national 
institutional contexts which affect unemployment risks in early career stages, but obviously 
also the broader structural context. The role of aggregate macroeconomic conditions, 
measured by either aggregate unemployment rates or employment growth rates, is a key 
determinant of unemployment risks among recent entrants to the labour market (van der 
Velden and Wolbers, 2000; Gangl, 2000b, 2000c). Those in their early career stages are 
particularly affected by cyclical market swings as they are typically among the less 
competitive individuals on the market and have not yet entered stable permanent job 
positions. In each recessionary period, unemployment rates among market entrants increase 
relatively stronger than aggregate rates, but they also decline more strongly in more buoyant 
times. As more detailed analyses show, it is the lowest qualified school leavers whose labour 
market chances are particularly vulnerable to cyclical macroeconomic developments (Gangl, 
2000c). But there is yet another important reason why those entering the labour market with 
low qualifications form a particular problem group. According to the results in Gangl (2000c), 
ongoing professionalisation of the labour force and related increases in skill requirements 
increasingly work against the lowest qualified school leavers. In addition, there is no evidence 
that any European country is exempted from this tendency. 

3.5.4 Types of jobs and the nature of employment contracts: some determinants 

Understanding unemployment risks is one important element in understanding transition 
processes, yet the flip side of the coin is to understand young people’s employment outcomes 
in their early career stages. Several of the project analyses have touched on these matters by 
addressing the nature of occupational allocation of labour market entrants (Gangl, 2000b, 
2000c) and the types of contracts obtained initially (van der Velden and Wolbers, 2000). We 
have not conducted any analysis on wage or earnings outcomes as no measures on them are 
provided in the EULFS, at least up to the 1997 wave which has been the most recent one 
considered in our work. And as in the case of unemployment, our main analytical interests 
focused on the role of education and training, institutions and socio-economic context factors 
in generating job outcomes at the start of individual careers. 




In fact, the role of education and training as a major individual resource in job competition 
emerges very clearly from our analyses. The higher the level of education attained, the higher 
the occupational status of job positions (Gangl, 2000b), the lower the likelihood of entering 
into low-skilled jobs (Gangl, 2000b), the higher the probability of accessing professional job 
positions already in the early career stage (Gangl, 2000b), the lower the probability of 
obtaining fixed-term or otherwise temporary job contracts (van der Velden and Wolbers, 
2000), and the higher the likelihood of having a full-time contract (van der Velden and 
Wolbers, 2000). And more specifically, it turns out that some of the particular advantages of 
apprenticeship contracts in terms of unemployment risks come at the expense of allocation to 
lower level jobs: compared to leavers from general or school-based vocational tracks at the 
upper secondary level, apprentices attain employment in lower status occupations and run a 
higher risk of entering low-skilled jobs. In fact, there are few indications that these 
relationships vary dramatically between the various European countries: in general, macro- 
institutional differences play a much more limited role with respect to employment outcomes 
among labour market entrants than is the case with respect to unemployment risks in the early 
career stages. 

Again, there is a certain role to play on the part of the institutional structure of education and 
training systems. In an almost trivial sense, the higher the level of educational attainment in a 
cohort entering the labour market, the higher will be the level of jobs for which they compete. 
The still lower educational levels in Southern Europe explain the, on average, lower 
occupational attainment levels in the early career stages there to a good deal already (Gangl, 
2000b). But potentially more interesting are the favourable effects of large scale dual systems 
or similar forms of vocationally specific training provision. According to our results, the 
presence of such systems lowers the incidence of low-skilled employment (Gangl, 2000b) and 
temporary contracts (van der Velden and Wolbers, 2000) among young people entering the 
labour market. To which extent this effect is due to the occupational specificity of the training 
provided itself or to the fact that the offered training tracks represent a low-threshold option 
for attaining education and training beyond compulsory levels is an open question to future 
research - but the effect itself is undeniably there. 


57 6 4 

Table 3.4 

Institutional Effects On Employment Outcomes among Market Entrants 

Study / 

Dependent Variable 

Institutional Variable 


van der Velden / 

Centralised Wage Bargaining 

negative, not significant 

Wolbers, 2000: 

Union Density 

no effect 

Temporary Contract 


Employment Protection 

Vocational Specificity / Educ. 

no effect 

Dual System 


Tracking / Second. Educ. System 

no effect 

van der Velden / 

Centralised Wage Bargaining 

positive, not significant 

Wolbers, 2000: 

Union Density 

no effect 


Employment Protection 

negative, not significant 

Vocational Specificity / Educ. 


Dual System 

negative, not significant 

Tracking / Second. Educ. System 

negative, not significant 

Gangl, 2000b: 

Apprenticeship Systems 

small negative effect 

Occupational Status 

Occupationalised Markets 

no effect 

Southern Europe 

no effect 

Gangl, 2000b: 

Apprenticeship Systems 

negative for secondary sector employment 



Occupationalized Markets 

positive for professional employment 

Southern Europe 

no effect 

Gangl, 2000c: 

Interaction of macroeconomic trends and 

positive effects of professionalisation 

Occupational Status 

three macro-institutional contexts 

strongest in OLM countries, negative effects 
of educational expansion strongest in 
Southern Europe 

Gangl, 2000c: 

Interaction of macroeconomic trends and 

positive effects of professionalisation 



three macro-institutional contexts 

strongest in OLM countries 

Apart from this, there is also some evidence for slightly different allocation mechanisms 
operating in the occupationalised markets of Austria, Denmark, Germany, and the 
Netherlands. In these systems, occupational allocation tends to be more strongly skill-based: 
as job competition relies more heavily on (formally certified) skills rather than experience, 
those entering the labour market in such contexts are relatively more competitive to adult 
workers than is the case in systems less reliant on certified skills. Hence, occupational and 
employment outcomes reflect more adequate matches at earlier career stages than elsewhere. 
In support of this reasoning, Gangl (2000b) provides evidence that higher levels of education 
provide more protection from entering low-skilled jobs in occupationalised systems, and that 
leavers from tertiary level education are much more likely to attain professional positions 
already at the outset of their careers. In addition, van der Velden and Wolbers (2000) 

establish an effect of the strictness of employment protection on labour market entrants' job 
outcomes. Paradoxically at first glance, the probability of obtaining initial employment on a 
fixed-term or temporary basis is higher in countries with stricter employment protection 
legislation. In fact, this might indicate a deliberate strategy to flexibilise youth labour markets 
so as to facilitate youth labour market integration, without at the same time sacrificing 
protection standards for the core work force (cf. Schroder, 2000). For the several other 
institutional indicators as tested in van der Velden and Wolbers (2000), results have not 
shown significant effects. 

In addition to these individual and institutional factors, the impact of macrostructural context 
factors is far from negligible. Actually, however, the role of aggregate macroeconomic 
conditions is much less important for occupational allocation and employment outcomes than 
for unemployment risks discussed earlier. At best, macroeconomic conditions determine only 
to a small part the extent to which those entering the labour market are allocated to lower 
level positions. In tighter labour markets, young people are disproportionately allocated into 
low-skilled and temporary jobs (Gangl, 2000b; van der Velden and Wolbers, 2000), and this 
allocation pattern is much less responsive to cyclical changes than are unemployment risks. 
Whether temporary jobs themselves are, in turn, more sensitive to the business cycle than 
permanent jobs, as they may have the role of a buffer to changes in product demand, is an 
open question for future research. 

But what turns out to be much more important to employment outcomes among labour market 
entrants is the (changing) balance between individual qualifications and skill demands on the 
market. Our analyses clearly show that net changes in the relative balance between supplied 
and demanded skills have important implications for employment outcomes among school 
leavers. In general, an increasing supply of better qualified market entrants triggers changes 
in allocation patterns at otherwise unchanged market conditions as better qualified leavers 
become substituted for less qualified ones. As a consequence, increasing levels of educational 
attainment have diminishing individual absolute and relative advantages as a by-product: on 
average, educational expansion implies lower occupational status outcomes, higher risks of 
low-skilled jobs, and decreasing probabilities of entering professional positions (Gangl, 
2000b, 2000c). In addition, educational expansion has also been accompanied by an 
expansion of part-time employment (van der Velden and Wolbers, 2000). Given the current 




trend of expansion at the tertiary level, the triggered adjustment reactions have, of course, 
mostly implied declining occupational returns among tertiary level leavers, and to a lesser 
extent, also among leavers from upper secondary education (Gangl, 2000c). As with many 
other results reviewed before, there is no indication in our data that these processes occur 
differently in different European countries. If anything, downward substitution pressures have 
even been somewhat stronger among tertiary level graduates in Southern European countries, 
potentially related to the strong ongoing catching-up processes in patterns of educational 

But as stated earlier, the net outcome of these developments is dependent on parallel changes 
in the structure of labour markets. To the extent that labour markets begin to utilise the higher 
level of supplied skills adequately, an increasing professionalisation of labour market demand 
actually counteracts the effects of educational expansion as young people in all European 
countries benefit from the increasing availability of employment positions appropriate to their 
skill levels (Gangl, 2000c). There is strong evidence that such labour market developments 
actually occurred, although probably somewhat time-lagged. That is, empirically, we do not 
observe particularly pervasive net changes in occupational outcome patterns over the past 
decades despite tremendous educational expansion because labour markets happened to 
generate increasing levels of demand for high-skill jobs. To understand if that correlation was 
purely incidental or whether both developments have in fact been closely interrelated and 
potentially intensified each other would appear as a pressing task for future research - not 
least in order to have clearer views on the policy implications of even further educational 
expansion. It might be that diminishing returns to education in the short run are an expression 
of short-run costs of adjustment to a modernised economic structure, which are in part borne 
by those entering the market in a period of restructuring. As this is still somewhat tentative, 
future research is clearly needed to provide answers on the nature of such driving forces 
behind changing patterns of occupational allocation. 

3.5.5 Summary 

In sum, the project’s analyses based on LFS data have stressed both the considerable 
similarity and also the substantial heterogeneity in European transition patterns. There are 
striking differences among EU member states in terms of the levels and types of qualifications 



which market entrants have at their disposal. There are important differences between the 
countries in terms of the institutional nature of training provision. There are excessively large 
differences between countries in terms of unemployment risks for those in their early career 
stages. And there are important differences in the types of jobs and the nature of employment 
contracts attained by young people. In fact, the project’s exploratory attempts to describe the 
variety of European transition patterns did not result in any definite picture of a European 
'map' of transition experiences. Still, it is probably fair to conclude that we have been able to 
bring out some contours more clearly than can be done on the basis of previous research: 
while the exceptional position of traditional dual system countries has already been the matter 
of much scholarly debate, notably in contrast to various other Northern European countries, 
the particular conditions applying in most Southern European countries have typically gone 
unnoticed in comparative research to date. 

Nevertheless, as shown by the more advanced analyses in the project, this heterogeneity of 
experiences does not necessarily defy systematic explanation. That is, the similarities and 
differences in European transition patterns described in this section can probably be explained 
as arising from some general underlying mechanisms which apply to all countries. In fact, for 
all the particular outcomes considered, we have compiled evidence of the importance of 
individual resources, notably education, institutional factors and broader socio-economic 
context conditions in generating the observed transition patterns. Typically, cross-national 
differences in the effects of any such resources and structural factors on youth labour market 
integration are quite small. For example, while there is some evidence for cross-national 
differences in the labour market value of (particular types of) education, the magnitude of 
such effects is often far from compelling. Similarly, there is little evidence that ongoing 
labour market changes affect young people in different ways in different European countries - 
what does differ between countries is the extent of macroeconomic turbulences rather than 
their effects on transition outcomes. 

Still, institutional factors often attain a prominent place in the explanation of cross-national 
similarities and differences. Country differences in macro-structural context conditions are 
usually of limited power in comparative explanations. In fact, a large part of cross-national 
variation turns out to be stable over time and cyclical economic changes, thus necessitating 
institutionally-based explanations. Among these, three particular institutional complexes 

figure prominently in our results: First, the institutional structure of education and training 
systems because this largely determines the nature of qualificational resources available to 
market entrants. Countries where young people achieve higher levels of education, as well as 
those countries operating large scale systems of vocational training, provide young people 
with a better start into working life. Second, the institutional labour market context, which 
governs the transformation of educational resources into employment outcomes. There are 
two aspects which have been addressed more extensively in project work, namely the role of 
occupational labour markets and the effects of employment protection legislation. It seems 
that occupationalised labour market contexts, that is, those labour markets tightly structured 
by occupational boundaries arising, for example, from the nature of educational supply and/or 
union action in recruitment processes, provide some advantages to young people as job 
competition relies more strongly on skills rather than experience. Hence, the relative 
competitiveness of young people is increased relative to systems more reliant on experience 
on the market. 

Finally, there is the issue of potential effects of employment protection legislation, which is 
often expected to negatively affect youth labour market integration. We have not found any 
evidence which would support this assumption. Rather, the evidence seems rather more 
consistent with a view that in more tightly regulated systems, the use of flexibly regulated 
forms of employment contracts (like fixed-term contracts or special forms of combined work- 
training contracts) is particularly widespread as a regulated means to foster the integration of 
young people into the labour market, rather than regulation amounting to a genuine 
impediment to integration itself. Countries like Italy and Greece are probably among those 
countries where such an argument is least likely to sound plausible, as tight employment 
protection standards are enforced, but provisions for flexible contracts to achieve youth 
integration are not really common, and both youth unemployment rates and the proportion of 
first-time job seekers among the unemployed are substantial. Still, we have little direct or 
indirect evidence for a destructive role of employment protection in the sense that employers 
appear particularly hesitant to recruit school leavers. In fact, two results we have obtained 
might be taken as indicative of the potentially crucial role of differences in supply-side 
behaviour for bringing about particular transition patterns in Southern Europe: empirically, 
what is specific about unemployment in, for example, Italy or Greece is the extent to which 
leavers from upper secondary, and even more so from tertiary, levels of education are 

affected. This is very much in contrast with what would have to be expected if employment 
protection were the main problem. In addition, there is some evidence that cyclical 
fluctuations in youth unemployment rates are considerably lower in most Southern European 
countries, which might also tentatively indicate a less dominant role of demand-side 
behaviour in shaping transition outcomes for young people. 


Section 3.5 has presented the main findings of analyses of Labour Force Survey data for 
fifteen European countries. In contrast, school leaver survey data are available for only five 
countries: France, Ireland, the Netherlands, Scotland and Sweden. While certain types of 
systems (in particular, dual system and Mediterranean countries) are excluded from this 
group, important differences related to distinct dimensions of educational differentiation and 
forms of labour market regulation are captured within the group. 

All of the systems can be regarded as highly standardised but differ in the extent and nature of 
differentiation within the same stage, and at the end of each stage, of education. The 
Netherlands has the most highly track-differentiated system with a distinction at both lower 
and upper secondary level between (different types of) academic and vocational courses. In 
France, there is a significant degree of tracking at upper secondary level, with different types 
of lycees and students studying for the BEP/CAP or different types of general or vocational 
Baccalaureat. Sweden represents an intermediate case, with over half of those at upper 
secondary level taking vocational programmes, albeit ones with a strong general component 
and little institutionalised linkage to the labour market. 

Ireland can be broadly characterised as a 'general' educational system, although track 
differentiation at upper secondary level has become increasingly apparent in recent years. 
Post-Leaving Certificate (PLC) vocational courses are provided within the school-based 
systems while two new programme options (the Vocational and Applied Programmes) have 
become available within the general upper secondary examination system. Scotland probably 
represents the clearest example of an undifferentiated school-based system, albeit with a 
number of students taking a mix of academic Highers and vocational modules. For some 
purposes, upper secondary provision in Scotland can be seen as encompassing a range of 

differentiated provision, including full-time school, Further Education (typically vocational or 
pre-vocational) and work-based training provision. However, because of the nature of the 
sample in the Scottish school leavers’ survey, for the purposes of this study we focus only on 
leavers from the school-based system, counting other forms of upper secondary provision as 
equivalent to early labour market destinations. 

The five countries also differ in the nature of formal differentiation at the end of each 
educational stage. Some systems (such as Ireland) have a highly differentiated grading 
structure with examination candidates awarded grades for individual subjects which may be 
taken at a number of curricular levels. Differentiated grading systems are also employed in 
Scotland, Sweden and the Netherlands. In contrast, systems, such as France and the 
Netherlands, differentiate only between 'passing' and 'failing' a particular stage. 

In sum, while school leaver survey data are unable to depict the whole range of education, 
training and labour market systems across Europe, they are nonetheless able to capture 
important dimensions of institutional variation in the transition process. In total, twelve 
working papers were prepared using the integrated school leavers' databases (see Table 3.5). 
These papers explored a range of topics, including cross-national variation in transition 
processes, participation in post-school training and differences among groups of young people 
in terms of gender and ethnicity. The following sections outline the main findings of these 
papers in terms of (i) educational outcomes, (ii) the relationship between education and the 
labour market, (iii) post-school training, and (iv) the social structuration of transition 





Table 3,5: Overview of SLS working papers 




Grelet, Y., Mansuy, 
M., Thomas, G. 

Transition from school to work and early 
labour force history 

Cross-national differences and 
similarities in the nature of 
transition processes 

Grelet, Y., Mansuy, 
M., Thomas, G. 

The transition process: towards exclusion 
or financial sufficiency, a French-Irish 

Prevalence of non-employment 
and low pay over the first five 
years in the labour market 

Hartkamp, J. and 
Rutjes, H. (2000a) 

A route to skills: a comparative analysis 
of the position of apprenticeship in 
transition systems in France, Ireland, the 
Netherlands and Scotland 

The level and nature of post- 
school apprenticeship 
participation across countries 

Hartkamp, J. and 
Rutjes, H. (2000b) 

Apprenticeship in Ireland, the 
Netherlands and Scotland: comparison of 
trends 1979-1997 

The level and nature of post- 
school apprenticeship 
participation across countries 
and over time 

Iannelli, C. (2000) 

School effects on youth transitions in 
Ireland, Scotland and the Netherlands 

Variation between schools 
within countries in post-school 
principal activity 

Iannelli, C. and Raffe, 
D. (2000) 

Vocational upper-secondary education 
and the transition from school to work 

Comparison of vocational and 
academic routes in Ireland, the 
Netherlands, Scotland and 

Mansuy, M. and 
Schroder, L. (2000) 

Immigrant youth in the labour market in 
France and Sweden 

Educational and labour market 
characteristics of immigrant 
youth in the two countries 

McCoy, S. (2000a) 

Relative labour market disadvantage 
amongst the least qualified in Ireland, 
Scotland, the Netherlands, France and 

Cross-national variation in the 
labour market position of the 
least qualified 

McCoy, S. (2000b) 

Relative labour market disadvantage 
among the least qualified: Ireland, the 
Netherlands and Scotland, 1979-1997 

Variation across countries and 
over time in the labour market 
position of the least qualified 

Schroder, L. (2000) 

The role of youth programmes in the 
transition from school to work 

Level and nature of 
participation in youth 
programmes across countries 

Smyth, E. (2000a) 

Gender differentiation in education and 
transition outcomes 

Cross-national differences and 
similarities in education and 
transition outcomes among 
women and men 

Smyth, E. (2000b) 

Gender differentiation in education and 
early labour market outcomes over time: 
a comparative analysis 

Variation across countries and 
over time in education and 
transition outcomes among 
women and men 





3.6.1 Educational outcomes 

Certain measures of educational attainment, such as CASMIN and ISCED, had been 
commonly used in previous cross-national studies. However, for our purposes these measures 
proved problematic. Firstly, school-leavers in our samples were not necessarily at the end of 
their education/training career. Secondly, these measures often ignored some of the 
dimensions of education which are conceptually and empirically important, at least in certain 
national contexts (such as examination grades). For this reason, we derived several 
dimensions of educational outcomes which we could use to capture the full complexity of 
cross-national variation; these included age on leaving school, educational level 
(incorporating stage and qualifications achieved), curricular track, grades received and 
subjects/courses taken (see CATEWE, 1999; Brannen, Smyth, 2000). 

Figure 3.9: Cross-national variation in educational level (1995/7) 

S? 20 

Dropout Lower level leaving 

Source: Calculated from McCoy (2000a). 

It was hypothesised that the institutional nature of the educational system would influence the 
proportion of young people exiting at different stages of their schooling. More specifically, it 
was expected that more general educational systems (such as Ireland) would have a higher 
proportion of less qualified leavers due to their comparative failure to retain those less 
academically oriented (McCoy, 2000a). While significant cross-national differences were 
apparent, this hypothesis was not wholly confirmed. Two distinct measures of lower 
qualifications were tested: 'drop-out', exiting the school system without any qualifications, 
and 'lower level leaving', exiting the school system without attempting upper secondary 
qualifications, a measure that also included 'drop-out'. The nature of cross-national variation 
depended on the precise measure of educational level used (see Figure 3.9). Rates of 'drop- 

■ Ireland 
H Netherlands 
B Scotland 
E Sweden 

out' (no qualifications) in the late 1990s were higher in Scotland than in Sweden, Ireland or 
the Netherlands 5 . If lower level leaving (no upper secondary qualifications) is considered, 
cross-national differences persist but the gap in exit rates between Scotland and the 
Netherlands/Ireland is somewhat reduced in magnitude (McCoy, 2000a). Lower level leaving 
remains much less prevalent in Sweden than in the other countries. Achieving an upper 
secondary qualification is highest in Sweden and Ireland and lowest in Scotland. It should be 
noted that these patterns refer to initial school-based education and do not incorporate other 
routes to upper secondary qualifications (such as apprenticeships or further education 
courses). The focus on school-based educational attainment has different implications for 
different countries; in Scotland, for example, many young people obtain upper secondary 
qualifications in post-school further education colleges (see Martin and Raffe, 1998). Cross- 
national differences in educational attainment are not static over time, however (Muller and 
Wolbers, 1999). Among the SLS countries on which data are available, Ireland and Scotland 
have both experienced a substantial growth over the 1980s and 1990s in the proportions 
staying in school to upper secondary level. In contrast, educational attainment levels in the 
Netherlands have remained relatively static, albeit over a shorter time-period (1989 to 1997) 
(McCoy, 2000b; Smyth, 2000b). 

Young people in the countries studied differ not only in their level of education but also in the 
type of education they receive. Differentiation between 'academic' and 'vocational' tracks is 
clear-cut within the Dutch and French contexts. In Sweden, vocational specialisations at 
upper secondary level were introduced in 1970 with reforms in the 1990s resulting in a 
reduction in the number of vocational programmes. In the Irish and Scottish cases, however, 
no such formal tracking exists within the general secondary school system, although students 
can differ markedly in the subjects they take. Furthermore, the Irish system has seen an 
increasing incorporation of vocational education (most commonly, Post-Leaving Certificate 
courses) into the upper secondary level. For our purposes, this distinction between the courses 
taken in Ireland and Scotland was seen as 'informal' tracking and the use of this concept 
allowed us to explore whether this 'informal' tracking operates in a similar fashion to more 
formally differentiated tracking. Given the institutional differences in the role of vocational 
education, it is hardly surprising that those taking vocational tracks represent a larger group in 

5 Rates of drop-out were also high in the French pattern. However, this pattern should be interpreted with some 
caution due to the exclusion of general Baccalaureat candidates from the survey sample. The French pattern 
alters markedly when this is taken into account (see Martin and Raffe, 1998). 



n a 


the Netherlands, Sweden and France (Iannelli and Raffe, 2000). Participation in vocational 
tracks has remained fairly stable in the Netherlands over the period 1989 to 1997. However, 
there has been a significant increase in the proportion of young Irish people taking vocational 
courses, reflecting the introduction and expansion of Post-Leaving Certificate provision over 
the period in question 6 (Smyth, 2000b). 

Within the countries considered, educational outcomes vary by gender, ethnicity and socio- 
economic background; this variation is considered in section 3.6.4 below. 

3.6.2 Education and the labour market 

One of the central concerns of the CATEWE project has been to examine the relationship 
between education and transition outcomes across a range of institutional and labour market 
contexts. A number of measures of labour market outcomes, such as the EGP social class 
schema, had previously been developed to examine stratification and mobility processes 
among the adult population. However, such measures caused difficulties when applied to our 
particular samples, leading to a high concentration of school-leavers in a small number of 
social class categories. Therefore, several dimensions of labour market outcomes were 
derived, including full/part-time status, the nature of the employment contract, social class, 
occupational status, occupational segment, industrial sector, industrial segment and earnings. 
The countries vary markedly along these dimensions with quite different industrial, 
occupational and earnings structures. A consideration of the factors underlying these cross- 
national differences lies outside the parameters of our study. Instead, analyses of the 
integrated SLS datasets have focused on exploring the way in which educational outcomes 
help to shape (variation in) young people's experiences of early transition processes. 



6 It is more difficult to examine trends over time in Scotland since the availability of data on the take-up of 
vocational modules varies over time. 


7 $ 

Figure 3 JO: A summary of the relationship between education and transition outcomes 

' outcome ' 

Educational outcome 



Educational type 


School variation 
(controlling for 




(except Ireland -; 

Scotland + at 
upper secondary 

(Ireland, Sweden, 









(Ireland, Scotland, 


(France 7 , 







State training 









Not significant 









sector job 



Part-time job 












Notes: + statistically significant positive relationship 

- statistically significant negative relationship 
n.s. no significant relationship 

shaded area - relationship not considered in the working papers 

It should be noted that in France young people can either transfer to apprenticeship programmes within their 
initial education (often at fairly early ages) or return to such programmes after leaving school. This finding 
relates to post-school participation. 



The presence in the integrated databases of information on distinct dimensions of education 
allowed us to explore the relative importance of particular educational outcomes in different 
national contexts. Analyses centred on four of these dimensions: differentiation in terms of 
educational level (stage left school or a combination of stage and qualifications received), 
between academic and vocational tracks, within stages in terms of examination grades, and 
between schools. The relationships between these different dimensions of educational 
background and a number of transition 'outcomes' are summarised in Figure 3.10. 

It was hypothesised that educational outcomes would have a significant influence on 
transition processes among young people in all of the study countries, due to the standardised 
nature of the qualifications systems considered. However, it was expected that in more 
'general' systems (such as Ireland and Scotland) educational level and grades received would 
assume a more important role in shaping transition outcomes while type of education 
(whether academic or vocational) would be more important in track-differentiated systems 
like the Netherlands (Hannan et al., 1999). 

Educational level 

The stage at which young people left school along with the qualifications they achieved were 
found to have significant influences on transition outcomes in all of the countries considered. 
Less qualified leavers are more likely to be unemployed than those with higher qualification 
levels (McCoy, 2000a) and tend to have longer spells of unemployment (Grelet et al., 2000a). 
Data from Ireland and France indicate that those without qualifications continue to be at a 
disadvantage in access to employment, even five years after entering the labour market 
(Grelet et al., 2000b). Less qualified leavers are also less likely to secure access to further 
education than those with upper secondary qualifications (Iannelli and Raffe, 2000; Smyth, 
2000b). In addition, the type and quality of job are associated with initial level of education; 
those with upper secondary qualifications have access to better quality jobs while the least 
qualified are more likely to have part-time jobs and receive low wages (Grelet et al., 2000a; 
McCoy, 2000a). Lower level leavers tend to be over-represented in manual employment and 
under-represented in the routine non-manual or professional classes. They are more likely to 
be found in secondary sector employment within the manufacturing or construction sectors 
and are less likely to secure employment in the finance, public administration or professional 
service sectors (McCoy, 2000a). 

While there are definite similarities across countries in the position of less qualified young 
people, their relative disadvantage tends to differ across countries and in terms of the labour 
market outcome considered. Unemployment risks are more strongly differentiated by initial 
educational level in Scotland than in Ireland, Sweden or the Netherlands (McCoy, 2000a). 
Within France, unemployment risks continue to be differentiated by educational level over the 
first five years in the labour market (Grelet et al., 2000b). The distribution of unemployment 
after the initial period of labour market entry also varies cross-nationally; in the Irish context, 
unemployment is concentrated within a small group who experience longer term 
unemployment while in the French context, unemployment is experienced by a broader group 
but interspersed with periods of short-term employment and training programmes (Grelet et 
al., 2000b). In terms of occupational allocation, access to professional employment is more 
strongly differentiated by education in Ireland and Scotland than in the the Netherlands or 
France. Furthermore, the relative disadvantage of lower level leavers in entry to secondary 
sector jobs is strongest in the Netherlands (McCoy, 2000a). 

There has been much debate about the growth of overqualification in the youth labour market. 
While such studies usually address labour market entrants from both secondary and tertiary 
levels (see, for example, Hannan et al., 1998), it might be expected that increasing educational 
levels coupled with growing or volatile unemployment rates would have some implications 
for changes in the returns to education among secondary leavers. No consistent picture of 
changes in educational returns over time emerges; however, there are tentative suggestions of 
declining returns in some countries and for certain labour market outcomes. Both the 
Netherlands and Scotland have experienced a decline in the gap between the most and the 
least qualified in their relative unemployment risks over time (McCoy, 2000b) with upper 
secondary leaving providing diminishing protection against unemployment in Scotland 
(Smyth, 2000b). There is also some evidence of declining differentials between educational 
levels in occupational status in the Netherlands and in secondary sector employment in 
Scotland and Ireland (McCoy, 2000b). 

Type of education 

The type of education received (whether academic or vocational) is considered in a number of 
analyses. It was hypothesised that the existence of formal track differentiation would mean 





that type of education would have the strongest effects in the Dutch, French and, to a lesser 
extent, Swedish systems. However, it was also recognised that 'informal' tracking may play a 
role in shaping transition outcomes in Ireland and Scotland. Differentiation in the tracks 
young people take through the secondary school system is, on average, highly predictive of 
their trajectories on leaving school. Those who take academic courses are much more likely to 
enter further education than those who have taken vocational courses. This difference is more 
marked in Netherlands and, perhaps surprisingly, Ireland than in Scotland 8 (Iannelli and 
Raffe, 2000). Thus, in the Irish and Dutch cases, school-based vocational courses provide 
occupationally-specific skills and therefore act as an alternative to acquiring such skills 
through further education. Analyses of the LFS data indicate that vocational education tends 
to be associated with a lower unemployment risk (see above). However, this pattern was not 
apparent in analyses of the school leavers' survey data, except in Ireland where young people 
who have taken vocational courses have lower unemployment rates than other groups 
(Iannelli and Raffe, 2000; McCoy, 2000a). It is likely that the differences between the two 
data sources can be accounted for by differences in the nature of the samples (apprentices 
were an 'outcome' for SLS but 'leavers' for LFS purposes), in the definition of vocational 
education (with a broader definition adopted in the SLS case) and the fact that unemployment 
rates in the highly track-differentiated Dutch case were extremely low at the time-point 

The type of education received was also found to have consequences for the nature of 
employment achieved. Having taken a vocational track increases the likelihood of entering 
manual employment, particularly skilled manual jobs, and appears to provide some protection 
against entry to secondary sector jobs in France, Ireland, the Netherlands, Scotland and 
Sweden (McCoy, 2000a). Conversely, those who have taken a vocational track are much less 
likely than academic leavers to enter professional employment or to find work in the 
distribution, finance or public administration sectors (McCoy, 2000a). It was hypothesised 
that acquiring occupationally-specific skills through school-based vocational education would 
also have a return in terms of pay levels. This hypothesis was confirmed with vocational 
leavers found to be at an earnings advantage in the more track-differentiated systems of the 
Netherlands and France (McCoy, 2000a). 

Unfortunately, France could not be included in these analyses. As general Baccalaureat leavers were not 
included in the sample, a full comparison of academic and vocational tracks at the upper secondary level could 
not be undertaken. 



The distinction between vocational and academic tracks appears to be an important one in 
shaping young people's early labour market experiences. In many ways, informal tracking in 
the Irish context appears to operate in a somewhat similar fashion to more formal tracking in 
the Netherlands and France. This pattern should be interpreted with some caution, however, 
since the effect primarily relates to participation in Post-Leaving Certificate courses which, in 
many ways, are more advanced in content as well as more occupationally specific than 
regular upper secondary courses (see Iannelli and Raffe, 2000). It is also important to go 
beyond a simple academic/vocational dichotomy to examine the type of vocational courses 
taken by young people. One such approach is to examine the gender composition of different 
vocational tracks 9 (see Smyth, 2000a; Smyth, 2000b). Many of the consequences of 
vocational education relate primarily to participation in the type of tracks usually dominated 
by young men, which increases the likelihood of entering skilled manual employment in the 
manufacturing or construction sectors. However, participants in mixed or female-dominated 
tracks are more likely to work in personal services or other non-manual employment (Smyth, 
2000a; 2000b). In this way, educational segregation plays a role in reproducing gender 
segregation within the labour market, although segregation is still apparent among young men 
and women who take the same kinds of vocational courses (see section 3.6.4 below). 

Examination grades 

The third dimension considered in analyses of the SLS databases related to differentiation 
within stages through examination grades. It was hypothesised that grades would have a more 
significant effect on transition outcomes in more general education systems than in more 
track-differentiated systems. This hypothesis was confirmed by the analyses. Among upper 
secondary leavers, grades are associated with further education entry in Scotland, Ireland and 
Sweden. Interestingly, grades are also associated with further education entry in the 
Netherlands. The latter pattern is likely to relate to some form of unmeasured heterogeneity 
among academic leavers (e.g. greater interest in further education among higher-performing 
students) since access to third-level education is not usually based on a numerus clausus 
system as it is in the Irish case. Grades also appear to be used by employers in making 
recruitment decisions since higher-performing students have reduced unemployment risks in 

9 More research is needed on the relationship between field of education and the nature of the transition process. 
Analysing the gender mix of vocational tracks is a way of assessing the relationship between educational and 
occupational segregation. However, it would also be useful to know if certain types of course content (e.g.